{"id":"fbd70cbb-867e-4b6c-91db-05467a6d5708","arxiv_id":"2501.08993","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Add(M) lies in Prod(M) exactly for Sigma-pure-injective objects under a non-omega-measurable cardinality hypothesis, and Prod(M)=Add(M) exactly for Sigma-pure-injective product-rigid objects, in two broad categorical settings.","lead":"This paper gives conditions under which the classes Add(M) and Prod(M) coincide or contain one another in locally finitely presented categories and in compactly generated triangulated categories. It also gives a positive answer to a version of Enochs's conjecture for classes of the form Prod(M) when no omega-measurable cardinals exist.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3.6 invokes Proposition 2.6 with an index set I whose non-omega-measurability is never established; the main implication therefore rests on an unsupported set-theoretic hypothesis.","rationale":"The reader's verdict is CONDITIONAL, and my analysis agrees that the paper should not be accepted as fully proved. The reader identified Lemma 2.4 as the weakest assumption, but I see the more acute problem in Theorem 3.6's application of Proposition 2.6: the index set I arising from the split epimorphism M^I -> M^(J) is not shown to satisfy Proposition 2.6's non-omega-measurability hypothesis. This is exactly the kind of missing justification that a conditional acceptance should demand. The reader's rationale does mention the same gap, so there is partial agreement; however, the reader's formal 'weakest assumption' points at Lemma 2.4 itself rather than at the unsupported applicability of that lemma. I am not claiming the theorem is false; rather, the written proof has a concrete hole that cannot be patched by the theorem's stated hypothesis without a new argument. The proposed compression lemma is the natural analytical check: if it holds, the proof can be repaired; if it fails, the theorem needs additional set-theoretic hypotheses. The paper's dependence on non-omega-measurable cardinals is explicit and legitimate, but the proof must ensure every index set to which Proposition 2.6 is applied is actually non-omega-measurable. No circularity or ad hominem concern is present; the issue is purely structural in the proof of Theorem 3.6.","tokens_in":14493,"tokens_out":16987,"duration_ms":182407,"concrete_test":"Attempt to prove the missing compression lemma: for every split epimorphism f: M^I -> M^(J) with |J| > |Hom(X,M)|, there exists a subset I' of I of non-omega-measurable cardinality such that the restriction f|: M^(I') -> M^(J) is again a split epimorphism. If this lemma cannot be proved, Theorem 3.6 remains unproved and the hypothesis must be strengthened, e.g. by assuming that the index sets I are non-omega-measurable or that no omega-measurable cardinals exist, as is done in Corollaries 3.10 and 3.11. A successful proof would close the gap; a counterexample would show the theorem as stated is false.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 3.6(i)->(ii) is the central argument of the paper. It assumes Add(M) subset Prod(M), fixes X in fpA and a descending chain of subgroups of finite definition, and chooses J with |J| > |Hom(X,M)|. From Add(M) subset Prod(M) it obtains a split epimorphism f: prod_{i in I} M -> oplus_{j in J} M. It then applies Proposition 2.6 to the Chase system with domain index set I. But Proposition 2.6 requires, as hypothesis (i), that I have non-omega-measurable cardinality. The theorem only assumes that Hom(X,M) has non-omega-measurable cardinality for every X in fpA. Nothing in the proof controls the set I: it is an arbitrary index set over which the product representing oplus_J M is taken, and no argument shows that I can be chosen non-omega-measurable, nor that I is bounded by |Hom(X,M)|. Since Lemma 2.4, and therefore Proposition 2.6, is explicitly false for omega-measurable index sets (Remark 2.5), the engine that forces the chain of subgroups of finite definition to become stationary may simply not be available. The reader's rationale noted this gap, but the stated weakest assumption was the induction inside Lemma 2.4. The more directly load-bearing issue is the unverified applicability of that lemma to the specific I produced by Add(M) subset Prod(M). Without a separate compression argument for I, Theorem 3.6 is not proved as stated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper characterizes the inclusions Add(M) ⊆ Prod(M) and Prod(M) ⊆ Add(M) in locally finitely presented additive categories and in compactly generated triangulated categories. The main results are Theorem 3.6 and Theorem 4.4, which equate Add(M) ⊆ Prod(M) with Σ-pure-injectivity under the hypothesis that all relevant Hom-groups have non-omega-measurable cardinality; Corollaries 3.10 and 4.8 apply this to the question of when Prod(M) is a (pre)covering class, and Corollaries 3.14 and 4.12 give the dual statements for Add(M). Theorems 3.13 and 4.10 characterize product-complete objects. The technical engine is a generalized Chase-type lemma (Lemma 2.4) and a consequence, Proposition 2.6, which force descending chains of subgroups of finite definition to become stationary.","tokens_in":14715,"tokens_out":13474,"duration_ms":145157,"significance":"The results are significant: they extend classical module-theoretic characterizations of Σ-pure-injectivity and product-completeness to a categorical setting, and they connect these notions to approximation theory and to Enochs-type conjectures. The paper is careful about the set-theoretic hypotheses, and Remarks 3.7 and 4.5 explicitly acknowledge that the main implications cannot be proved in ZFC if omega-measurable cardinals are allowed. The categorical framework via the evaluation functor and the purity category is well chosen, and the transfer to compactly generated triangulated categories is a natural and useful extension. I found no circularity: the proofs build on the paper's Chase-type lemma together with cited external results. The main reservation concerns a load-bearing missing verification in the proof of Theorem 3.6, described below.","major_comments":[{"comment":"The proof applies Proposition 2.6 to the Chase system indexed by the set I obtained from Add(M) ⊆ Prod(M), but Proposition 2.6 requires, as hypothesis (i), that I have non-omega-measurable cardinality. The only non-measurability assumption in Theorem 3.6 concerns the cardinalities of Hom(X,M) for X ∈ fpA; nothing in the proof controls the index set I. Indeed, Add(M) ⊆ Prod(M) asserts only that the direct sum ⊕_{j∈J} M is a direct summand of some product ∏_{i∈I} M, and the set I is otherwise arbitrary. No argument is given that I can be chosen non-omega-measurable, nor that |I| is bounded in terms of |Hom(X,M)|. Since Remark 2.5 records that Lemma 2.4 and hence Proposition 2.6 fail for omega-measurable index sets, the conclusion that the chain (Hom(Y_k,M)b_k) is stationary is not justified as written. This gap is load-bearing for Theorem 3.6 and propagates to Theorem 4.4 via the application of Theorem 3.6. The proof should either show that the existence of a split epimorphism ∏_{i∈I} M → ⊕_{j∈J} M with |J| > |Hom(X,M)| forces the existence of a non-omega-measurable I satisfying the same hypotheses, or modify the argument so that Proposition 2.6 is applied to an index set whose non-measurability is actually verified.","section":"Theorem 3.6, proof of (i)⇒(ii), Section 3.3"}],"minor_comments":[{"comment":"The statement says \"If there exists a subset I in the power-set of I\"; it should say \"If there exists a family \\mathcal{I} of subsets of I\". The overloaded notation I for both the set and the family makes the proof harder to read.","section":"Lemma 2.3, Section 2"},{"comment":"The transfer from a Prod(ev(M))-cover to a Prod(M)-cover is compressed: a cover K is a direct summand of ev(M^I) (for some product), not of ev(X), so it would be clearer to state explicitly that Lemma 3.9 is applied with A = M^I.","section":"Corollary 3.10, Section 3.3"},{"comment":"In the proof of (iii)⇒(i), the sentence \"Since for A is Σ-pure-injective\" should read \"Since M is Σ-pure-injective\".","section":"Theorem 3.13, Section 3.4"},{"comment":"The step \"From [12, Theorem 2.11], it follows that for every finitely presented object K from Mod-T^c the group Hom(K,HM) is not omega-measurable\" relies on a nontrivial transfer from representable functors to all finitely presented functors; a brief explanation of how the cited theorem applies would improve readability.","section":"Theorem 4.4, Section 4.2"}],"recommendation":"major_revision","confidential_remarks":"The central results are plausible and well motivated, but the proof of Theorem 3.6 has a genuine gap concerning the non-omega-measurability of the index set I. If this can be repaired, the paper would be a solid contribution. The issue is significant enough that I cannot recommend acceptance in the current form."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe short version: this paper has a nice new Chase-type lemma and a sensible extension of known module-theoretic results to locally finitely presented and compactly generated triangulated categories, but the main theorem as written rests on an unverified set-theoretic hypothesis. It deserves a serious referee, but the authors need to fill a gap before the result is established.\n\nWhat's actually new: Lemma 2.4 is a genuinely useful generalization of Chase's lemma to non-omega-measurable index sets, proved by a clean induction using Lemma 2.3. The characterizations in Theorem 3.6 and Theorem 4.4 (Add(M) ⊆ Prod(M) iff M is Sigma-pure-injective, under non-omega-measurable Hom sets) extend known module-level results to broader categories, and the applications to covering/enveloping classes (Corollaries 3.10–3.14 and 4.8–4.12) are well motivated. The paper is honest about the set-theoretic dependence, noting Remark 3.7 and citing Saroch's counterexample. I saw no circularity.\n\nThe soft spot is exactly where the stress-test note points. In Theorem 3.6(i)⇒(ii), the proof takes Add(M) ⊆ Prod(M), fixes X and a descending chain of subgroups of finite definition, and from a split epimorphism f: ∏_{i∈I} M → ⊕_{j∈J} M obtains a Chase system. It then applies Proposition 2.6, which requires I to be non-omega-measurable. But nothing in the proof shows that the I produced by the split epimorphism can be chosen with non-omega-measurable cardinality. The theorem's hypothesis only bounds Hom(X,M), not I. The reader's note flagged this too, and the stress-test note is right that this is the load-bearing issue. It looks fixable—one might enlarge I or chase through the construction—but as written the implication is not proved. The same issue might affect the triangulated analogue, Theorem 4.4.\n\nAlso, Corollary 4.8 asserts finite generation of the right ideal Ker(Hom(M^J, α)) without a clear argument; that's a less central concern but worth tightening.\n\nWho is this for: anyone working on purity, Sigma-pure-injectivity, or approximation theory in additive or triangulated categories. If the gap is repaired, this would be a solid contribution.\n\nMy recommendation: send it to peer review, with a referee who knows the set-theoretic background. The chase lemma and the applications are valuable enough that a good referee can help the authors fix the index-set issue. Worth engaging.","headline":"The paper has a genuinely useful new Chase-type lemma and a plausible transfer of the Add/Prod characterizations to locally finitely presented and triangulated categories, but Theorem 3.6 as written has a real, likely fixable gap: the index set I is never shown to be non-omega-measurable before invoking Proposition 2.6.","tokens_in":15329,"tokens_out":1967,"would_cite":false,"duration_ms":19378,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18G80","18A25","18G05","20K25"],"pacs":[],"model":"deepseek-v4-flash","headline":"Add(M)⊆Prod(M) iff M is Σ-pure-injective, in locally finitely presented and compactly generated triangulated categories, when Hom-sets are non-ω-measurable; Prod(M) is then (pre)covering and covering-closure follows.","keywords":["locally finitely presented categories","Σ-pure-injective objects","product-complete objects","Chase's Lemma","precovering classes","preenveloping classes","compactly generated triangulated categories","ω-measurable cardinals"],"falsifier":"Search for an object $M$ in a locally finitely presented category such that $|\\mathrm{Hom}(X,M)|$ is non-$\\omega$-measurable for every finitely presented $X$, $\\mathrm{Add}(M)\\subseteq\\mathrm{Prod}(M)$ holds, yet some $\\mathrm{Hom}(X,M)$ has a strictly descending chain of subgroups of finite definition. By Theorem 3.4 that chain is exactly the failure of $\\Sigma$-pure-injectivity, so its existence would refute Theorem 3.6; the proof's own Proposition 2.6 would be the step forced to fail.","tokens_in":14203,"feed_emoji":"🔗","tokens_out":20574,"duration_ms":167052,"temperature":0.7,"pith_summary":"The paper asks when the two closure classes generated by a single object $M$—direct summands of direct sums, $\\mathrm{Add}(M)$, and direct summands of direct products, $\\mathrm{Prod}(M)$—contain one another. Its main theorem says that in a locally finitely presented additive category, if $|\\mathrm{Hom}(X,M)|$ is not $\\omega$-measurable for every finitely presented $X$, then $\\mathrm{Add}(M)\\subseteq\\mathrm{Prod}(M)$ holds exactly when $M$ is $\\Sigma$-pure-injective. The same characterization is proved for compactly generated triangulated categories via the restricted Yoneda functor. Because $\\Sigma$-pure-injectivity also forces $\\mathrm{Prod}(M)$ to be closed under direct sums and pure quotients, the equivalence turns into concrete approximation-theoretic consequences: $\\mathrm{Prod}(M)$ is a (pre)covering class precisely in that case, and, assuming there are no $\\omega$-measurable cardinals, such covering classes are closed under directed limits.","feed_headline":"Add(M) fits inside Prod(M) exactly when M is Σ-pure-injective","feed_subtitle":"A generalized Chase lemma transfers the module equivalence to locally finitely presented and triangulated categories.","key_machinery":"The engine is the generalized Chase lemma, Lemma 2.4. A Chase system is a homomorphism $\\varphi:\\prod_{i\\in I}A_i\\to\\bigoplus_{j\\in J}B_j$ of abelian groups together with descending chains of subgroups $A_{i,k}$ and $B_{j,k}$ such that $\\varphi$ sends $\\prod_i A_{i,k}$ into $\\bigoplus_j B_{j,k}$; the lemma asserts that if $I$ is not $\\omega$-measurable, then $\\varphi$ maps the product of sufficiently deep subgroups over all but finitely many coordinates into a finite direct sum plus the intersection over $n$ of the direct sums of the $B_{j,n}$. Proposition 2.6 converts this into a descending-chain stationarity result under surjectivity and cardinality hypotheses, and Theorem 3.4 identifies that stationarity with $\\Sigma$-pure-injectivity. In the triangulated half, the restricted Yoneda functor $H:T\\to\\mathrm{Mod}\\text{-}T^c$ moves the question into a locally finitely presented Grothendieck category, where the same machinery applies.","core_discovery":"On the paper's own terms, the central discovery is Theorem 3.6: for an object $M$ in a locally finitely presented category $\\mathcal{A}$ with $|\\mathrm{Hom}(X,M)|$ non-$\\omega$-measurable for every finitely presented $X$, the inclusion $\\mathrm{Add}(M)\\subseteq\\mathrm{Prod}(M)$ is equivalent to $M$ being $\\Sigma$-pure-injective. The forward direction formalizes a split epimorphism $M^I\\to M^{(J)}$ as a Chase system and applies the generalized Chase lemma to show that all descending chains of subgroups of finite definition in $\\mathrm{Hom}(X,M)$ are stationary, which is the standard certificate of $\\Sigma$-pure-injectivity. The reverse direction observes that $\\Sigma$-pure-injectivity makes the summation map from a direct sum of copies of $M$ into a product a split monomorphism. Theorem 4.4 repeats the equivalence in compactly generated triangulated categories, and Corollaries 3.10 and 4.8 convert it into the statement that $\\mathrm{Prod}(M)$ is precovering, covering, or definable exactly under the same condition.","pith_inferences":["A reusable consequence the paper leaves implicit is that any proof built on Chase's lemma over countable index sets can be rerun over non-$\\omega$-measurable index sets whenever the Hom-groups involved have non-$\\omega$-measurable cardinality, which broadens the range of module-theoretic arguments that transfer to larger products.","The set-theoretic boundary is sharp: because a known construction under large-cardinal assumptions produces a non-$\\Sigma$-pure-injective free module with $\\mathrm{Add}(F)\\subseteq\\mathrm{Prod}(F)$, the theorem's equivalence cannot be a theorem of ZFC; the non-$\\omega$-measurability hypothesis is what keeps the proof inside ZFC.","For objects $M$ whose Hom-sets $\\mathrm{Hom}(X,M)$ are all countable—or more generally below the first measurable cardinal—the characterization applies unconditionally, making the $\\mathrm{Add}(M)\\subseteq\\mathrm{Prod}(M)$ condition a directly checkable test for $\\Sigma$-pure-injectivity in many concrete categories."],"forward_implications":["If every $\\mathrm{Hom}(X,M)$ has non-$\\omega$-measurable cardinality, then $\\mathrm{Prod}(M)$ is a precovering class exactly when it is a covering class, and exactly when $M$ is $\\Sigma$-pure-injective.","Assuming no $\\omega$-measurable cardinals exist, any precovering class of the form $\\mathrm{Prod}(M)$ is closed under directed limits, a positive case of the conjecture that every covering class is closed under directed limits.","The equality $\\mathrm{Prod}(M)=\\mathrm{Add}(M)$ holds exactly when $M$ is $\\Sigma$-pure-injective and product-rigid (every local-endomorphism-ring direct summand of a product of copies of $M$ is a direct summand of $M$); consequently $M$ is product-complete exactly when $\\mathrm{Add}(M)$ is a (pre)enveloping class.","In algebraic compactly generated triangulated categories, $\\mathrm{Prod}(M)$ is definable, precovering, or covering exactly when $M$ is $\\Sigma$-pure-injective, giving the triangulated analogue of the approximation consequences."],"supporting_citations":[{"why":"Provides the original Chase lemma for countably indexed products that Lemma 2.1 generalizes.","marker":"[9]"},{"why":"Supplies the extension to iterated direct sums and products and the counterexample showing the lemma fails for ω-measurable index sets.","marker":"[11]"},{"why":"Gives the module-level argument and the descending-chain algorithm that Proposition 2.6 and Theorem 3.6 adapt.","marker":"[7]"},{"why":"Provides the background on locally finitely presented categories, purity, subgroups of finite definition, and the characterization of Σ-pure-injective objects used as Theorem 3.4.","marker":"[19]"},{"why":"Constructs a free module with Add(F)⊆Prod(F) that is not Σ-pure-injective under large-cardinal assumptions, showing the non-ω-measurability hypothesis is needed.","marker":"[27]"},{"why":"Gives the set-theoretic facts about ω-measurable cardinals behind Lemma 2.3 and the cardinality hypothesis.","marker":"[12]"},{"why":"Supplies the definability and approximation results that turn Σ-pure-injectivity into precovering and covering conclusions.","marker":"[24]"},{"why":"Provides the characterization of Σ-pure-injectivity in compactly generated triangulated categories used in Theorem 4.2.","marker":"[5]"},{"why":"Establishes the classical module-theoretic link between product-complete objects and enveloping classes that Corollaries 3.14 and 4.12 extend.","marker":"[1]"}],"fun_headline_variants":["Add(M) ⊂ Prod(M) iff M is Σ-pure-injective","Σ-pure-injectivity: the exact condition for Add(M) ⊆ Prod(M)","When does Add(M) fit into Prod(M)? Only if M is Σ-pure-injective","The exact link: Add(M) ⊆ Prod(M) ⇔ M is Σ-pure-injective"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The characterization rests on the generalized Chase lemma's claim that a homomorphism from a product of abelian groups over a non-$\\omega$-measurable index set into a direct sum almost factors through finitely many coordinates; if that set-theoretic factorization fails, the proof that $\\mathrm{Add}(M)\\subseteq\\mathrm{Prod}(M)$ forces $\\Sigma$-pure-injectivity collapses.","fun_headline_variants_meta":{"raw":{"variants":["Add(M) ⊂ Prod(M) iff M is Σ-pure-injective","Σ-pure-injectivity: the exact condition for Add(M) ⊆ Prod(M)","When does Add(M) fit into Prod(M)? Only if M is Σ-pure-injective","The exact link: Add(M) ⊆ Prod(M) ⇔ M is Σ-pure-injective"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001177,"raw_usage":{"total_tokens":4819,"prompt_tokens":858,"completion_tokens":3961,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":474,"completion_tokens_details":{"reasoning_tokens":3866}},"tokens_in":474,"tokens_out":3961,"duration_ms":25575,"temperature":1.0,"reasoning_tokens":3866,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:13:25.480039+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for an object $M$ in a locally finitely presented category such that $|\\mathrm{Hom}(X,M)|$ is non-$\\omega$-measurable for every finitely presented $X$, $\\mathrm{Add}(M)\\subseteq\\mathrm{Prod}(M)$ holds, yet some $\\mathrm{Hom}(X,M)$ has a strictly descending chain of subgroups of finite definition. By Theorem 3.4 that chain is exactly the failure of $\\Sigma$-pure-injectivity, so its existence would refute Theorem 3.6; the proof's own Proposition 2.6 would be the step forced to fail.","supporting_citations":[{"cited_title":"Chase, On direct sums and products of modules , Pacific J","cited_arxiv_id":null,"evidence_quote":"Provides the original Chase lemma for countably indexed products that Lemma 2.1 generalizes."},{"cited_title":"Dugas, B","cited_arxiv_id":null,"evidence_quote":"Supplies the extension to iterated direct sums and products and the counterexample showing the lemma fails for ω-measurable index sets."},{"cited_title":"Breaz, Σ -pure injectivity and Brown Representability , Proc","cited_arxiv_id":null,"evidence_quote":"Gives the module-level argument and the descending-chain algorithm that Proposition 2.6 and Theorem 3.6 adapt."},{"cited_title":"Krause, Homological Theory of Representations , Cambridge University Press, 2022","cited_arxiv_id":null,"evidence_quote":"Provides the background on locally finitely presented categories, purity, subgroups of finite definition, and the characterization of Σ-pure-injective objects used as Theorem 3.4."},{"cited_title":"ˇSaroch, Σ -algebraically compact modules and Lω 1ω -compact cardinals, Math","cited_arxiv_id":null,"evidence_quote":"Constructs a free module with Add(F)⊆Prod(F) that is not Σ-pure-injective under large-cardinal assumptions, showing the non-ω-measurability hypothesis is needed."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the set-theoretic facts about ω-measurable cardinals behind Lemma 2.3 and the cardinality hypothesis."},{"cited_title":"Laking and J","cited_arxiv_id":null,"evidence_quote":"Supplies the definability and approximation results that turn Σ-pure-injectivity into precovering and covering conclusions."},{"cited_title":"Bennett-Tennenhaus, Characterisations of Σ -pure-injectivity in triangulated categories and applications to endoperfect objects , Fund","cited_arxiv_id":null,"evidence_quote":"Provides the characterization of Σ-pure-injectivity in compactly generated triangulated categories used in Theorem 4.2."},{"cited_title":"Angeleri-H¨ ugel,Covers and envelopes via endoproperties of modules , Proc","cited_arxiv_id":null,"evidence_quote":"Establishes the classical module-theoretic link between product-complete objects and enveloping classes that Corollaries 3.14 and 4.12 extend."}],"review_version":1}