{"id":"918eb12d-0e90-4b91-8205-ffa650f5043f","arxiv_id":"2411.17975","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Cotorsion pairs are defined in (d+2)-angulated categories, characterized geometrically in type A cluster categories, and shown to be stable under mutation.","lead":"This paper defines cotorsion pairs, a way of pairing subcategories inside higher-dimensional triangulated categories, and shows these pairs survive a process called mutation. It also gives a geometric rule for detecting these pairs in cluster categories of type A, which are used in representation theory.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The key reduction in Theorem 4.11 conflates the subfactor shift T^d with Σ^d: a U-angle T^{-d}X -> B -> ... -> X is not automatically a C-angle Σ^{-d}X -> B -> ... -> X, so the lifting step and hence Theorem 4.13 need an additional argument.","rationale":"The reader's weakest assumption focuses on the hypotheses on D and Z failing; my concern is internal to the proof even when those hypotheses hold. Both point to Theorem 4.11/4.13, but mine is the more load-bearing issue because it affects the central reduction to the subfactor category. The paper's examples and the overall strategy are plausible, and the gap may be fixable by adding a careful argument that translates standard angles in U back to C-angles with the correct first object and intermediate terms. Since the reader already gave a conditional accept, my verdict remains unchanged: the authors should be asked to expand the proof of Theorem 4.11, especially the surjectivity direction, and to fix the typo in Definition 3.7 noted by the reader. I do not see grounds for rejection, as no counterexample or contradiction is established; the concern is about completeness and rigor of the central argument.","tokens_in":30,"tokens_out":31198,"duration_ms":311094,"concrete_test":"Independently re-derive the surjectivity step of Theorem 4.11 for d = 2, writing out the standard-angle construction from Lemma 2.3 and all rotations explicitly. Concretely, in the 4-angulated category OA_2^2 of Example 3.16, take a nonzero D satisfying the hypotheses, such as D = add(135) (after verifying the hypotheses), and a cotorsion pair whose core contains D, e.g., (X23, Y23) with X23 = Y23 = {135, 136, 146}. Check whether the U-angle T^{-2}X -> B -> Y1 -> Y2 -> X lifts to a C-angle of the form Σ^{-2}X' -> C -> Y1' -> Y2' -> X' with X' ∈ X and Y_i' ∈ Y. If the lifted angle starts at T^{-2}X rather than Σ^{-2}X, the proof has a real gap.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central mutation theorem (Theorem 4.13) rests on Theorem 4.11, which asserts a bijection between cotorsion pairs in C whose cores contain D and cotorsion pairs in U = Z/D. The surjectivity proof of Theorem 4.11 has an underjustified step. Starting from a cotorsion pair (Xbar, Ybar) in U and an object B ∈ Z, one obtains a U-angle T^{-d}X -> B -> Y1 -> ... -> Yd -> X. The text then says: 'This implies a (d+2)-angle in C Σ^{-d}X -> B -> Y1 -> ... -> Yd -> X.' This is exactly the load-bearing step. The shift T^d in U is defined through D-mutation (Lemma 4.8 and Section 2.2), and Lemma 4.9(2) only gives an isomorphism of Hom-spaces U(X, T^dY) ≅ C(X, Σ^dY) for X,Y ∈ Z; it does not identify the object T^{-d}X with Σ^{-d}X in C. A standard angle in U with first object T^{-d}X is represented by a C-angle starting at T^{-d}X, not automatically at Σ^{-d}X. Moreover, the angle produced by Lemma 2.3 from the displayed diagram is Y1 -> Y2⊕D1 -> ... -> X⊕Dd -> Σ^dC -> Σ^dY1, which requires several rotations and shifts to match either form in Definition 3.1(2); those steps are omitted and it is not apparent that the intermediate objects land in X and Y in the required pattern. Without this missing argument, the bijection and hence Theorem 4.13 are unsupported.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper defines cotorsion pairs and weak cotorsion pairs in (d+2)-angulated categories, generalizing Nakaoka's cotorsion pairs in triangulated categories. It proves basic closure properties, gives a geometric characterization of weak cotorsion pairs in (d+2)-angulated cluster categories of type A using the non-intertwining relation, and then aims to prove that mutation of a cotorsion pair (and, according to the abstract, of a weak cotorsion pair) is again a cotorsion pair. The mutation theorem is approached through a bijection between cotorsion pairs in C whose cores contain a subcategory D and cotorsion pairs in the subfactor category U = Z/D, where Z = ⊥(Σ^dD) = (Σ^{-d}D)^⊥.","tokens_in":17194,"tokens_out":17150,"duration_ms":137678,"significance":"If the main theorem is correct, it would extend the known mutation-invariance of cotorsion pairs in triangulated categories (Zhou–Zhu) to the higher homological setting, and the geometric characterization of weak cotorsion pairs in type A provides explicit examples and a clear combinatorial description. The paper also gives useful examples and connects the new definitions to cluster tilting and Oppermann–Thomas cluster tilting objects. However, the central mutation theorem is currently supported by a proof with a load-bearing gap, and the abstract claims a result for weak cotorsion pairs that is not proved in the body, so the significance is contingent on a substantial revision.","major_comments":[{"comment":"Definition 3.7(2) states 'For any C ∈ X there exists a (d+2)-angle Xd → ... → X0 → C → Σ^dXd with Xi ∈ X'. This must be 'For any C ∈ C', since cluster tilting is a condition on all objects of the ambient category. As printed, Lemma 3.8 is false: the cotorsion pair (X,X) gives approximations for every object of C, not only for objects of X. This is a definitional error that affects the subsequent use of cluster tilting subcategories in Example 3.9 and in the statement of Lemma 3.8.","section":"Definition 3.7 and Lemma 3.8"},{"comment":"The proof asserts that a (d+2)-angle in U of the form T^{-d}X → B → Y1 → ... → Yd → X 'implies a (d+2)-angle in C' of the form Σ^{-d}X → B → Y1 → ... → Yd → X. This step is not justified. Lemma 4.9(2) only provides an isomorphism of Hom-spaces U(X, T^dY) ≅ C(X, Σ^dY); it does not identify the object T^{-d}X with Σ^{-d}X in C. A standard U-angle starting at T^{-d}X is, by construction, represented by a C-angle starting at T^{-d}X, not at Σ^{-d}X. The subsequent diagram and the application of Lemma 2.3 require a full morphism of (d+2)-angles, but the text only supplies the first and last vertical maps, and the vanishing C(Σ^{-d}X, D)=0 does not produce the needed intermediate vertical morphisms. Since this step is the bridge from cotorsion pairs in U to cotorsion pairs in C, the bijection asserted in Theorem 4.11 and hence Theorem 4.13 are unsupported as written.","section":"Theorem 4.11, proof of surjectivity"},{"comment":"The abstract and introduction claim that 'any mutation of a (weak) cotorsion pair in C is again a (weak) cotorsion pair'. Section 4 proves this only for cotorsion pairs (Theorem 4.13). No theorem or argument is given for mutation of weak cotorsion pairs; the geometric characterization in Theorem 3.14 is not used to establish such a statement. Thus the central claim in the abstract is broader than the results actually proved.","section":"Abstract, Introduction, and Section 4"}],"minor_comments":[{"comment":"The proof of Lemma 3.8 is only 'It is easy to check by the definitions'; given the misprint in Definition 3.7, a short proof or a precise reference for the equivalence would be helpful.","section":"Lemma 3.8"},{"comment":"The assertion 'The cotorsion pairs we can find are only (C,0) and (0,C)' is stated without proof; a verification or a reference would improve the example.","section":"Example 3.10"},{"comment":"The phrase 'which is denoted by Z' after the equality ⊥(Σ^dD) = (Σ^{-d}D)^⊥ is ambiguous: it is the common value of the two subcategories that is denoted by Z, not D itself.","section":"Theorem 4.13"},{"comment":"The notation for the image of a cotorsion pair in U is not distinguished from the original pair in C; the text writes (X,Y) for both. Introducing an overline or a different font would prevent confusion in the bijection statement.","section":"Theorem 4.11"},{"comment":"There are several typos and spacing issues, e.g., 'concide' in Remark 3.6 and the title page 'P AIRS' / 'T RA TES' artifacts; these should be corrected in the final version.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The gap in Theorem 4.11 is not a minor omission: it is the only argument connecting cotorsion pairs in the subfactor category to cotorsion pairs in the original category. If the authors cannot supply a lemma showing that T^{-d}X can be replaced by Σ^{-d}X in the relevant C-angles (or an alternative direct proof of Theorem 4.13), the main mutation theorem is not established. The abstract's claim about weak cotorsion pairs should either be proved or removed from the abstract and introduction. The paper is within the scope of the journal, and the geometric characterization part is sound, but the central theorem needs substantial reworking."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the paper does something useful: it carries cotorsion pairs and weak cotorsion pairs from triangulated categories into (d+2)-angulated categories, proves a geometric characterization in type A, and states a mutation theorem for d≥2. That is a natural next step after Zhou–Zhu and Holm–Jørgensen–Rubey, and the definitions are sensible. But there is a real gap in the central mutation result, and a typo that makes Lemma 3.8 false as printed.\n\nDefinition 3.7 says cluster tilting requires an angle “for any C ∈ X” where it should be “for any C ∈ C.” As printed, Lemma 3.8 (“(X,X) cotorsion iff X cluster tilting”) is false, because the “if” direction would only cover objects of X. This is an obvious typo, but it needs fixing.\n\nMore concerning is the surjectivity step in Theorem 4.11. Starting from a cotorsion pair in the subfactor category U, the proof obtains a U-angle T^{-d}X → B → Y1 → ⋯ → Yd → X, then asserts “This implies a (d+2)-angle in C Σ^{-d}X → B → ⋯ → X.” That does not follow. The shift T^d in U is defined by D-mutation, not by Σ^d; Lemma 4.9 gives an isomorphism of Hom-spaces U(X,T^dY) ≅ C(X,Σ^dY), but it does not identify the objects T^{-d}X and Σ^{-d}X in C. The underlying C-angle represented by a U-angle starting at T^{-d}X starts at T^{-d}X, not at Σ^{-d}X. Without an additional argument showing these can be exchanged, the lifting step and Theorem 4.13 are unsupported. The subsequent diagram chase is fine once that step is supplied.\n\nEverything else looks plausible. The geometric characterization in Section 3.2 is a straightforward translation of the orthogonality condition via Oppermann–Thomas’ Lemma 3.12—mildly tautological, but it gives a clean statement and the worked example in Example 3.16 is useful. The paper is honest about citing its d=1 templates.\n\nBottom line: this is a paper a serious referee should see. The definitions and the mutation theorem, if the gap can be closed, are worth having. The typo is trivial; the gap in 4.11 is the main thing to send back.\n\nRecommendation: send to peer review, conditional on the authors repairing the 4.11 step.","headline":"A useful higher-angulated extension with a genuine gap in the mutation theorem: the lift from the subfactor category in Theorem 4.11 swaps T^{-d}X for Σ^{-d}X without justification.","tokens_in":17723,"tokens_out":10875,"would_cite":true,"duration_ms":87610,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18E40","05E10","18G80"],"pacs":[],"model":"deepseek-v4-flash","headline":"A new notion of cotorsion pair is defined for (d+2)-angulated categories, and any mutation of such a pair is shown to remain a cotorsion pair.","keywords":["cotorsion pair","weak cotorsion pair","(d+2)-angulated category","mutation","cluster category","cluster tilting subcategory","higher homological algebra"],"falsifier":"Compute, in a (d+2)-angulated category where Z is not extension closed, a cotorsion pair (X, Y) with D ⊆ I(X) satisfying the perp equality, and verify that $μ_D^{{-1}}$(X) and $μ_D^{{-1}}$(Y) fail to satisfy the Hom-vanishing or approximation condition required of a cotorsion pair; alternatively, in OA_d^n, exhibit a diagonal set X with X ≠ nc nc X and show directly via Lemma 3.12 that (X, nc X) is not a weak cotorsion pair.","tokens_in":16601,"feed_emoji":"🔺","tokens_out":5632,"duration_ms":41896,"temperature":0.7,"pith_summary":"The paper introduces cotorsion pairs and weak cotorsion pairs in (d+2)-angulated categories, extending the classical triangulated-category notion (d=1). Its central goal is to show that this structure behaves well under mutation: starting from a cotorsion pair whose core contains a suitable subcategory D, the forward and backward D-mutations of both components again form a cotorsion pair, and the core transforms by the same mutation. As an application, the authors give a complete geometric description of weak cotorsion pairs in (d+2)-angulated cluster categories of type A: they are exactly pairs of diagonal sets X, Y with X = nc Y and Y = nc X, where nc is the non-intertwining complement. A sympathetic reader would care because this supplies a higher-homological analogue of the torsion/cotorsion pair mutation theory that underpins cluster algebra structures, and it predicts that mutation is a symmetry of the cotorsion-pair poset.","feed_headline":"Any mutation of a cotorsion pair is again a cotorsion pair","feed_subtitle":"This gives a geometric classification of weak cotorsion pairs in type A cluster categories.","key_machinery":"The carrying mechanism is the subfactor category U = Z/D associated to a D-mutation pair (Z, Z), where morphisms are taken modulo those factoring through D; the quotient inherits a (d+2)-angulated structure when Z is extension closed. Inside U, the functor U(X, T^d Y) is isomorphic to C(X, Σ^d Y), which lets the authors transfer cotorsion pairs between C and U. The geometric classification in type A is carried by the intertwining relation ≀ on diagonals and its complement nc X: non-vanishing of Hom(O_i, O_j[d]) is exactly the intertwining of the two diagonals, so the orthogonal conditions defining weak cotorsion pairs become X = nc Y and Y = nc X.","core_discovery":"The main structural theorem (Theorem 4.13) states: if (X, Y) is a cotorsion pair in a (d+2)-angulated category C and D is a strongly functorially finite subcategory contained in the core I(X) = X ∩ Y, satisfying ⊥(Σ^d D) = ($Σ^{{-d}}$ D)^⊥ and with Z = ⊥(Σ^d D) extension closed, then both ($μ_D^{{-1}}$(X), $μ_D^{{-1}}$(Y)) and (μ_D(X), μ_D(Y)) are cotorsion pairs, and I($μ_D^{{-1}}$(X)) = $μ_D^{{-1}}$(I(X)), I(μ_D(X)) = μ_D(I(X)). The proof reduces the D-mutation to a 0-mutation (the d-suspension) inside the subfactor category U = Z/D, which is itself (d+2)-angulated; the compatibility between cores in C and in U is what transfers the property back. For d = 1 this recovers the classical mutation result for torsion pairs in triangulated categories.","pith_inferences":["The bijection between cotorsion pairs in C containing D and all cotorsion pairs in the subfactor category U suggests that iterated mutations can be composed and understood as moves in a lattice of cotorsion pairs, analogous to exchange graphs in cluster theory; the paper does not explicitly develop this lattice perspective.","The geometric type-A classification likely extends to other finite (d+2)-angulated categories with Calabi-Yau properties: the nc-complement structure depends only on the intertwining dimension-vector formula, so one could test whether every finite (d+2)-angulated cluster category admits an analogous diagonal model.","A testable consequence: in type A, the equation X = nc nc X should characterize the first component of a weak cotorsion pair; one could enumerate all diagonal sets for small n and d and check that the failure of this equation exactly predicts failure of the weak cotorsion condition."],"forward_implications":["Any mutation of a cotorsion pair (respectively weak cotorsion pair) in a (d+2)-angulated category is again a cotorsion pair (respectively weak cotorsion pair), so mutation acts as a symmetry on the poset of cotorsion pairs.","The core of a cotorsion pair is preserved up to the same mutation: I(μ_D^{-1}(X)) = μ_D^{-1}(I(X)) and I(μ_D(X)) = μ_D(I(X)).","In (d+2)-angulated cluster categories of type A, weak cotorsion pairs are in bijection with pairs of non-intertwining diagonal sets satisfying X = nc Y and Y = nc X, giving a purely combinatorial classification.","For d = 1, the theorem specializes to the known mutation result for torsion pairs in triangulated categories, so the paper's framework is a genuine higher-dimensional generalization.","Since (X, X) is a cotorsion pair exactly when X is cluster tilting, the mutation result includes the statement that mutations of cluster tilting subcategories are again cluster tilting in this setting."],"supporting_citations":[{"why":"Supplies the definition and basic axioms of (d+2)-angulated categories and the standard construction from d-cluster tilting subcategories.","marker":"[8]"},{"why":"Provides D-mutation pairs and the construction of the subfactor (d+2)-angulated quotient category Z/D used throughout Section 4.","marker":"[21]"},{"why":"Supplies the (d+2)-angulated cluster categories OA_d^n, the diagonal model, and the intertwining criterion for non-vanishing Hom-spaces (Lemma 3.12).","marker":"[24]"},{"why":"The d=1 triangulated result that mutation of torsion pairs preserves torsion pairs, which the paper generalizes.","marker":"[28]"},{"why":"The mutation theory in triangulated categories that motivates the definition of mutation pairs.","marker":"[12]"},{"why":"The Ptolemy-diagram/nc-complement characterization of torsion pairs in type A cluster categories, recovered as Corollary 3.15 when d=1.","marker":"[10]"}],"fun_headline_variants":["Mutation keeps cotorsion pairs intact in (d+2)-angulated categories","Higher angulated categories: cotorsion pairs persist under mutation","Cotorsion pairs mutate to cotorsion pairs in any dimension","Weak cotorsion pairs: mutation invariance and type A geometry","New proof: mutation preserves cotorsion pairs in higher categories"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The mutation theorem rests on the assumption that the mutating subcategory D is strongly functorially finite and d-rigid with ⊥(Σ^d D) = ($Σ^{{-d}}$ D)^⊥ and that Z = ⊥(Σ^d D) is extension closed; if extension-closedness fails, the quotient category may not carry the (d+2)-angulated structure on which the proof depends.","fun_headline_variants_meta":{"raw":{"variants":["Mutation keeps cotorsion pairs intact in (d+2)-angulated categories","Higher angulated categories: cotorsion pairs persist under mutation","Cotorsion pairs mutate to cotorsion pairs in any dimension","Weak cotorsion pairs: mutation invariance and type A geometry","New proof: mutation preserves cotorsion pairs in higher categories"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000882,"raw_usage":{"total_tokens":3782,"prompt_tokens":889,"completion_tokens":2893,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":505,"completion_tokens_details":{"reasoning_tokens":2802}},"tokens_in":505,"tokens_out":2893,"duration_ms":17988,"temperature":1.0,"reasoning_tokens":2802,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:40:24.969191+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute, in a (d+2)-angulated category where Z is not extension closed, a cotorsion pair (X, Y) with D ⊆ I(X) satisfying the perp equality, and verify that $μ_D^{{-1}}$(X) and $μ_D^{{-1}}$(Y) fail to satisfy the Hom-vanishing or approximation condition required of a cotorsion pair; alternatively, in OA_d^n, exhibit a diagonal set X with X ≠ nc nc X and show directly via Lemma 3.12 that (X, nc X) is not a weak cotorsion pair.","supporting_citations":[{"cited_title":"Geiss, B","cited_arxiv_id":null,"evidence_quote":"Supplies the definition and basic axioms of (d+2)-angulated categories and the standard construction from d-cluster tilting subcategories."},{"cited_title":"Lin, n-angulated quotient categories induced by mutation pairs","cited_arxiv_id":null,"evidence_quote":"Provides D-mutation pairs and the construction of the subfactor (d+2)-angulated quotient category Z/D used throughout Section 4."},{"cited_title":"Oppermann, H","cited_arxiv_id":null,"evidence_quote":"Supplies the (d+2)-angulated cluster categories OA_d^n, the diagonal model, and the intertwining criterion for non-vanishing Hom-spaces (Lemma 3.12)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The d=1 triangulated result that mutation of torsion pairs preserves torsion pairs, which the paper generalizes."},{"cited_title":"Iyama, Y","cited_arxiv_id":null,"evidence_quote":"The mutation theory in triangulated categories that motivates the definition of mutation pairs."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The Ptolemy-diagram/nc-complement characterization of torsion pairs in type A cluster categories, recovered as Corollary 3.15 when d=1."}],"review_version":1}