{"id":"8e3a5e45-94a3-4ee8-8beb-e9bbc153fa26","arxiv_id":"2607.15991","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A finite-rank abelian length category is brick-finite if and only if it is widely determined and widely co-determined.","lead":"An abelian category with finitely many simple objects is brick-finite exactly when every torsion class is generated by a wide subcategory and every torsionfree class is cogenerated by a wide subcategory. This gives a uniform, lattice-theoretic criterion that extends known results for finite-dimensional algebras to all finite-rank length categories.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unproved assertion in Lemma 4.2 — infinite semibrick generates non-compact torsion class — leaves a load-bearing gap in Prop 4.3.","rationale":"The reader's weakest assumptions identified both the unproved Lemma 4.2 assertion and the reliance on Theorem 3.9's exchange bijection. My stress test confirms that Lemma 4.2 is the more load-bearing gap: it is an internal assertion, not a citation, and it is essential for the finite branching argument in Proposition 4.3. The rest of the proof — Lemma 5.1, Lemma 5.4, and the Zorn argument — appears internally consistent, and the cited Asai–Pfeifer decomposition is plausibly valid in the stated generality. Since the unproved assertion is likely true but not demonstrated, the correct verdict is CONDITIONAL: the paper is promising but needs a proof of the infinite-semibrick non-compactness claim (and preferably a fuller writing-out of the dual computations in Theorem 3.9).","tokens_in":14240,"tokens_out":46217,"duration_ms":454276,"concrete_test":"Provide a rigorous proof of Lemma 4.2: if an infinite semibrick B generates a compact torsion class T(B)=T(M), then every B∈B is a subquotient of a finite direct sum of M, so B must be finite due to the finite length of M. If the proof fails, test a possible counterexample using the Kronecker quiver's infinite family of regular simples of dimension 2; determine whether the torsion class they generate is finitely generated. This would settle whether the assertion holds in the required generality.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The proof of the main equivalence (Theorem 5.5) depends on Lemma 4.2, which asserts that 'for any given infinite semibrick B the torsion class T(B) is not compact.' This is used in Proposition 4.3 to conclude that the tree of saturated paths is finitely branching, and then König's lemma gives an infinite branch and a non-compact torsion class, a contradiction. The assertion is plausible: a compact torsion class is finitely generated by an object M of finite length, and an infinite semibrick should not fit inside T(M) because T(M) contains only finitely many isomorphism classes of bricks. But the paper gives no proof, and the step is not obvious in the full generality of abelian length categories (as opposed to finite-dimensional algebras, where it is standard). If the assertion were false, the finite branching argument collapses and Proposition 4.3 — and hence the converse direction of Theorem 5.5 — would not be established. This is a genuine gap in the written argument, but it is repairable, so the verdict should remain conditional rather than reject.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a lattice-theoretic characterisation of brick-finiteness for abelian length categories of finite rank: such a category is brick-finite if and only if every torsion class is generated by a wide subcategory and every torsionfree class is cogenerated by a wide subcategory (Theorem 1.1, equivalently Theorem 5.5). The proof introduces an exchange bijection between the upper/lower brick labels of two torsion pairs forming a wide interval (Theorem 3.9), uses it to show local finiteness of Hasse-neighbourhoods along saturated chains (Lemma 5.1), and then proves by a Zorn's lemma argument that every torsion class is reached by a finite saturated chain (Lemma 5.4). A König's lemma argument in Proposition 4.3 converts upper torsion connectedness into brick-finiteness. As a corollary, the first Brick–Brauer–Thrall theorem is extended to this setting.","tokens_in":14488,"tokens_out":36876,"duration_ms":405571,"significance":"If correct, the main theorem is a substantial conceptual advance: brick-finiteness, a representation-theoretic property, is shown to be equivalent to a purely lattice-theoretic property of the torsion-pair lattice. The paper also gives a direct proof of the exchange of brick labels across wide intervals, a phenomenon that the author describes as a shadow of simple-minded mutation, and derives the bounded-brick-length theorem. The arguments are mostly elementary and carefully organised, and the reliance on external results is explicitly signposted. The proof of the converse direction, however, relies on a nontrivial assertion in Lemma 4.2 that is neither proved nor referenced, which leaves a genuine gap in the written argument.","major_comments":[{"comment":"The proof of Lemma 4.2 reduces to the assertion that 'for any given infinite semibrick B the torsion class T(B) is not compact.' This assertion is load-bearing: Proposition 4.3 uses it to conclude that the tree of saturated paths is finitely branching, since the upper labels of a fixed torsion class form a semibrick, and then applies König's lemma to obtain a non-compact torsion class. No proof or reference is provided for the assertion. In the stated generality of abelian length categories it is not immediate: compactness of T(B) means T(B)=T(M) for a finite-length object M, and one must show that an infinite Hom-orthogonal set of bricks cannot be contained in the extension closure of quotients of a single finite-length object. For finite-dimensional algebras this follows from known τ-tilting/brick results, but the paper does not cite or prove an analogue for arbitrary abelian length ca","section":"Lemma 4.2 (p. 15)"}],"minor_comments":[{"comment":"The step 'C should be in the torsion class t' is terse. It would help to spell out that a proper quotient of a simple object of W lies in ⊥0 W, so that C, being a quotient of Q, indeed lies in t = u ∩ ⊥0 W.","section":"Proposition 3.3, Step 2"},{"comment":"The statement 't has exactly N neighbours in the Hasse quiver' should clarify whether this means the total number of upper and lower covers; for t=0, for instance, there are N upper labels and no lower labels.","section":"Lemma 5.1"},{"comment":"There are a few typographical issues ('first-Brick Brauer Thrall', 'the first-Brick Brauer Thrall theorem') and some sentences that would benefit from copy-editing, e.g. 'More recently, in the second version of the survey [23], appeared a characterisation...'.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The central idea and most of the proof are sound, and the gap in Lemma 4.2 appears repairable, either by a short argument exploiting finite composition length or by a reference to an existing theorem. I would be willing to recommend acceptance once the authors supply the missing justification. I do not see grounds for rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a genuinely new characterization of brick-finiteness for finite-rank abelian length categories, and the exchange bijection in Section 3 is the real workhorse. It deserves a proper referee, though there's one unproved step that needs fixing before I'd trust the converse direction.\n\nThe main theorem — brick-finite iff WD and WCD — extends known artin algebra results to a natural generality and recovers the first Brick–Brauer–Thrall theorem as a corollary. That is a solid contribution, not just a reformulation. The exchange bijection (Theorem 3.9) is new at this level of abstraction and the proof is mostly detailed: the envelope arguments in Propositions 3.3 and 3.6 are carefully done, and the forward direction of the main theorem is routine. The examples in Section 6 are genuinely useful, especially the k[x]-module and A-infinity illustrations.\n\nThe soft spots are real but repairable. The main one is Lemma 4.2: the claim that an infinite semibrick always generates a non-compact torsion class is asserted in a single sentence and used to get finite branching in Prop 4.3. This is not obvious in full generality — a compact torsion class can contain many bricks, and it is not immediate that an infinite Hom-orthogonal family cannot be absorbed into a finitely generated torsion class. If this assertion fails, the König's lemma argument collapses and the converse direction of Theorem 5.5 is unsupported. I suspect the claim is true, but the paper needs either a proof or a precise citation. The other gaps are minor: the dual half of Theorem 3.9 is dismissed with \"one checks\", and the proof leans heavily on Asai–Pfeifer's decomposition, but those are standard in the field and not problematic.\n\nOverall, the paper is well-written, honest about its debts, and the central argument is coherent. The gaps are patchable, not fatal. I'd send it to review with a request to expand Lemma 4.2 and fill in the dual computations. The result is worth having in the literature, and I would likely cite it in my own work.","headline":"Solid extension of brick-finiteness theory to finite-rank length categories, with a clean exchange bijection as the core new tool; one unproved compactness assertion in Lemma 4.2 needs patching before the converse direction is airtight.","tokens_in":14984,"tokens_out":16123,"would_cite":true,"duration_ms":172176,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["16G10","18E10","18E40"],"pacs":[],"model":"deepseek-v4-flash","headline":"Brick-finiteness of a length category is equivalent to every torsion class being generated and every torsionfree class cogenerated by wide subcategories.","keywords":["torsion pairs","bricks","wide subcategories","abelian length categories","brick-finiteness","lattice of torsion classes","Hasse quiver","brick length"],"falsifier":"Find a finite-rank abelian length category with a wide interval (t,f) ⊂ (u,v) for which t ≠ u∩⊥0(u∩f); then the exchange bijection cannot exist and the main theorem would not follow by this argument. Alternatively, exhibit a finite-rank WD and WCD category with infinitely many bricks, which would directly falsify the claimed equivalence.","tokens_in":14089,"feed_emoji":"🧱","tokens_out":6596,"duration_ms":54549,"temperature":0.7,"pith_summary":"Brick-finiteness—having only finitely many isomorphism classes of bricks—is a strong finiteness property for abelian length categories. This paper proves that, for categories with finitely many simple objects, brick-finiteness is equivalent to a purely lattice-theoretic condition: every torsion class must be generated by a wide subcategory, and every torsionfree class must be cogenerated by a wide subcategory. The proof passes through an exchange bijection that moves brick labels across a cover in the torsion-class lattice, a mechanism the author describes as a shadow of simple-minded collection mutation. If correct, the result recasts a representation-theoretic finiteness property as an order-theoretic one and extends the bounded-length criterion for brick-finiteness to all finite-rank length categories.","feed_headline":"Brick-finiteness equals wide generation and cogeneration","feed_subtitle":"Length categories are brick-finite precisely when every torsion class and torsionfree class come from a wide subcategory.","key_machinery":"The exchange bijection (Theorem 3.9). For a wide interval, i.e. a pair of torsion classes t ⊊ u such that the intersection W = u∩f is a wide subcategory, the theorem builds an explicit bijection between the upper and lower brick labels of u and those of t, using W-envelopes and W-covers. The bijection transfers the count of Hasse neighbours across a cover, and this count propagation is what forces local finiteness of the Hasse quiver along every saturated chain. A cited decomposition of wide intervals (t = u∩⊥0W, f = W∗v) supplies the structural basis for the exchange.","core_discovery":"The paper establishes that a finite-rank abelian length category is brick-finite if and only if it is both widely determined and widely co-determined: every torsion class t arises as T(W) for some wide subcategory W and every torsionfree class f arises as F(W′) for some wide subcategory W′. The equivalence is proved by showing, in one direction, that these two wide-generation conditions force every torsion class to lie at finite Hasse distance from zero, and then that finite distance plus finite branching implies finiteness of the torsion-pair lattice, hence brick-finiteness by the standard criterion. In the other direction, brick-finiteness immediately gives compactness and hence the two wi","pith_inferences":["If the equivalence holds, brick-finiteness is an invariant of the torsion-class lattice alone, so any two categories with isomorphic lattices would agree on brick-finiteness; this could be tested by finding lattice-isomorphic but representation-inequivalent examples.","The exchange bijection gives an explicit mutation recipe for brick labels across covers, which may extend to non-brick-finite categories as a mutation rule for semibrick pairs; a natural test is whether the bijection remains bijective without the wide generation and cogeneration hypotheses.","The paper's locally brick-finite condition for infinite-rank categories suggests WD+WCD may be strictly weaker than local brick-finiteness; constructing an infinite-rank WD and WCD category that is not locally brick-finite would settle whether the finite-rank theorem has a clean infinite counterpart.","In the artin algebra setting, a one-sided condition (widely co-determined) is known to suffice; comparing the minimal hypotheses needed in general length categories could point to a finer hierarchy of brick-finiteness-like properties."],"forward_implications":["Brick-finiteness can be checked by verifying wide generation and wide cogeneration, properties stated only in terms of the torsion-pair lattice.","In a brick-finite category with N simple objects, every torsion class has exactly N neighbours in the Hasse quiver of the torsion-class lattice; the Hasse diagram is regular.","Every torsion class in a brick-finite category is reachable from zero by a finite saturated chain, and the lattice of torsion pairs is finite.","Bounded brick length is equivalent to brick-finiteness for all finite-rank abelian length categories."],"fun_headline_variants":["Brick-finiteness iff wide generation and cogeneration","Wide subcategories determine brick-finiteness in length categories","Brick-finite length categories from wide torsion classes","Length categories: brick-finite iff wide torsion and cotorsion"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof leans on a decomposition of wide intervals (t = u∩⊥0W, f = W∗v) that is cited from earlier work rather than proved; if that decomposition fails in any finite-rank length category, the exchange bijection carrying the induction collapses.","fun_headline_variants_meta":{"raw":{"variants":["Brick-finiteness iff wide generation and cogeneration","Wide subcategories determine brick-finiteness in length categories","Brick-finite length categories from wide torsion classes","Length categories: brick-finite iff wide torsion and cotorsion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000266,"raw_usage":{"total_tokens":1379,"prompt_tokens":608,"completion_tokens":771,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":352,"completion_tokens_details":{"reasoning_tokens":702}},"tokens_in":352,"tokens_out":771,"duration_ms":7647,"temperature":1.0,"reasoning_tokens":702,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T21:41:46.737099+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a finite-rank abelian length category with a wide interval (t,f) ⊂ (u,v) for which t ≠ u∩⊥0(u∩f); then the exchange bijection cannot exist and the main theorem would not follow by this argument. Alternatively, exhibit a finite-rank WD and WCD category with infinitely many bricks, which would directly falsify the claimed equivalence.","supporting_citations":[],"review_version":1}