{"id":"4c1a2a43-59a0-46cb-93dd-d5c47f2ec38e","arxiv_id":"2411.17200","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"In semi-abelian varieties of algebras, all Yoneda n-step extension collections between fixed objects admit a set of representatives, so cohomology objects are small.","lead":"This paper proves that in semi-abelian varieties of algebras, the collection of n-step extensions between any two objects is small, meaning it admits a bijection to a set. It matters because Yoneda-style cohomology objects in non-additive settings depend on such extension collections being legitimate sets rather than proper classes.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the central smallness proof for n-step extensions in semi-abelian varieties is sound; Section 6's proof sketch is peripheral and does not affect Theorem 5.2.","rationale":"The reader's weakest_assumption points to the syzygy reduction and the use of Choice; I independently checked those steps and found them valid. The syzygy theorem relies only on standard pullback stability and existence of weakly universal normal epimorphisms, both present in semi-abelian varieties. The cardinality argument in Theorem 4.8 is sound: the collection of normal monos into subsets of K^ℓ × Q is a set, and the transported structure makes the resulting representative a genuine short exact sequence. The Choice-dependent fibre selection in Theorem 5.2 is a standard way to turn a surjection onto a conglomerate into a bijection with a set. The only caveat is Section 6, where Theorem 6.1 is given as a proof sketch; however, the paper's central claim, Theorem 5.2, does not depend on it. I therefore see no reason to change the reader's ACCEPT verdict.","tokens_in":15139,"tokens_out":32634,"duration_ms":296869,"concrete_test":"As a belt-and-braces check, write out the pullback construction of Theorem 5.1 explicitly for n = 2 in a non-abelian semi-abelian variety, verifying that the spliced pulled-back sequence admits a morphism of exact sequences to the original 2-step extension; if a square fails to commute, the syzygy reduction needs repair. Separately, expand the proof sketch of Theorem 6.1 into a full proof, since it is the only stated theorem not given a complete argument.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I examined the central argument in good faith and found no load-bearing flaw. The one-step bound (Theorem 4.8) is the key step: Corollary 4.6 produces a pointed-set retraction ψ: X → K^ℓ × Q over Q, and although ψ is not an algebra homomorphism, the transported algebra structure on the image X1 makes ψ an isomorphism of algebras. With this transported structure, k1 = ψk is a normal monomorphism and q1 = π_Q|X1 is a normal epimorphism, indeed the cokernel of k1 because ψ is an iso and q1ψ = q. The set NM(Q,K,ℓ) is small by powerset, replacement, and the fact that a set carries only a set of algebra structures. The syzygy reduction (Theorem 5.1) is the second load-bearing step: pullback-stability of normal epimorphisms and enough normal-projectives guarantee that the pulled-back sequence is exact and that splicing it with the syzygy yields a morphism to the original sequence, so w is surjective. The use of Choice to select fibre representatives is legitimate and non-circular. The only under-supported passages are Theorem 6.1 (explicitly a proof sketch) and Section 7's higher Schreier case (explicitly conditional on an open question); both are presented as variations, not as premises for Theorem 5.2. No internal inconsistency in the central proof surfaced.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves that in any semi-abelian variety of universal algebras, the conglomerate of all n-step Yoneda extensions between two fixed objects is small, i.e., in bijection with a set. The main argument is in two parts: Theorem 4.8 uses the Bourn–Janelidze characterisation of semi-abelian varieties to show that every one-step extension is represented by a normal monomorphism whose codomain is a subset of K^ℓ×Q, so that Ext^1(Q,K) is small; Theorems 5.1–5.2 then use a syzygy argument, assuming pullback-stable normal epimorphisms and enough normal-projectives, to reduce smallness of Ext^n to smallness of Ext^1. The paper also sketches variations for double and crossed extensions and treats Schreier extensions of monoids.","tokens_in":15321,"tokens_out":18734,"duration_ms":164193,"significance":"The result closes a genuine foundational gap: Yoneda-style cohomology objects in non-additive semi-abelian categories are not automatically known to have underlying sets, and standard homotopical-algebraic tools do not apply directly. The proof of the main theorem is elementary, self-contained, and gives explicit representatives; the Axiom of Choice is invoked transparently in the syzygy reduction. If accepted, the paper justifies the use of Yoneda Ext in semi-abelian varieties and provides a template for handling similar size questions elsewhere. The variations in Sections 6–7 are clearly marked as sketches or conditional results and do not affect the main theorem.","major_comments":[],"minor_comments":[{"comment":"The statement that a one-step extension is \"a retract over Q of the short exact sequence 0→K^ℓ→K^ℓ×Q→Q→0\" is stronger than what the proof establishes: the proof gives pointed-set maps φ and ψ with φψ=1_X, π_Q ψ=q and qφ=π_Q, i.e., a retraction of the middle object X onto a subset of K^ℓ×Q over Q, but it does not construct maps between the kernel objects K and K^ℓ making the diagram of sequences commute. Since the kernel objects differ, the sequence-level reading cannot be literally true; the corollary should be rephrased as an object-level retraction, which is all that Theorem 4.8 uses.","section":"§4, Corollary 4.6"},{"comment":"The smallness of double extensions is announced as a theorem but supported only by a proof sketch that relies on external material on regular pushouts and double syzygies; because this result is not used in the main theorem, this is acceptable, but the paper should make explicit at that point that a full proof is not included here.","section":"§6, Theorem 6.1"},{"comment":"The last paragraph correctly states that the smallness of Ext^n_S(Q,K) for n≥2 is conditional on the existence of a weakly universal Schreier extension for each monoid Q; it might help the reader if this conditional status were also recorded in the abstract or introduction.","section":"§7, final paragraph"},{"comment":"The caption contains a typo: \"Spicing\" should be \"Splicing\".","section":"Figure 1 caption"},{"comment":"There are a few missing spaces and inconsistent hyphenation (e.g., \"semiabelian\" vs. \"semi-abelian\", \"normal epimorphismq1\"); these should be corrected in the final copy.","section":"Throughout"}],"recommendation":"minor_revision","confidential_remarks":"The main theorem is sound and the proof is largely self-contained; I recommend minor revision. The only substantive issue is the overstatement in Corollary 4.6, which should be corrected even though it does not affect Theorem 4.8. The Section 6 material is appropriately labelled as a sketch but is considerably less detailed than the main sections."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper proves what its title promises: in any semi-abelian variety, the collection of n-step Yoneda extensions between two fixed objects is small, meaning it admits a bijection to a set. That is a genuinely useful foundational result. People have been quietly ignoring this size issue for decades, and this paper shows that in the varietal setting it is harmless. The proof is direct and, as far as I can tell, correct.\n\nWhat is new: the one-step smallness argument (Theorem 4.8) builds on the authors' own earlier retract lemma, but here it is proved in full and then extended to all n via a syzygy reduction (Theorem 5.2). The syzygy argument is standard but carefully done, with the use of Choice made explicit. The paper is honest about what it does not do: Section 6 on double extensions is presented as a proof sketch that leans on external results, and Section 7 on Schreier extensions is conditional on an open question. Those are peripheral to the main claim, and the authors flag them as variations, not as premises.\n\nThe paper does several things well. It states the set-theoretic foundations clearly, so the meaning of “small” is unambiguous. It gives a self-contained proof of the key one-step bound using the Bourn–Janelidze characterization of semi-abelian varieties, and it is careful about the distinction between the conglomerate of extensions and a chosen set of representatives. The citation pattern looks sensible: previous work is cited where it is used, and the overlap with the authors' own Corollary 4.6 in [35] is acknowledged while the new material is clearly the reduction to arbitrary length.\n\nWhere are the soft spots? Section 6 is the weakest part. Theorem 6.1 is stated with a proof sketch, and while I do not doubt the conclusion, a referee should push for either a full proof or an explicit statement that this part is provisional. Section 7 is openly conditional, so it is less of a problem. The paper is also modest in scope: it does not produce new cohomology computations or reorganize the field, but that is not a flaw. It settles a question that should have been settled long ago.\n\nWho should read it: anyone doing Yoneda-style cohomology in semi-abelian categories, and anyone who teaches or uses extensions in groups, Lie algebras, loops, or related structures. It deserves a serious referee. My recommendation: send it to review, and ask the authors to either expand the Section 6 sketch or mark it clearly as a separate conjecture. The central theorem is solid.","headline":"A clean, correct proof that Yoneda extension collections in semi-abelian varieties are small; worth a serious referee, though the double-extension part is only sketched.","tokens_in":15951,"tokens_out":1167,"would_cite":true,"duration_ms":12360,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["03E25","03E30","18E13","18G15","18G50"],"pacs":[],"model":"deepseek-v4-flash","headline":"In semi-abelian varieties of algebras, the collection of all n-step cohomology extensions between two objects is small—in bijection with a set of representatives.","keywords":["Yoneda extension","semi-abelian variety","semi-abelian category","smallness of extension classes","syzygy argument","Schreier extension","crossed extension","cohomology"],"falsifier":"Find a semi-abelian variety $\\mathcal{V}$ and objects $K,Q$ in $\\mathcal{V}$ for which the collection of isomorphism classes of short exact sequences $0 \\to K \\to X \\to Q \\to 0$ is a proper class, so that no set of representatives exists; this would refute Theorem 4.8 and, with it, the main smallness theorem.","tokens_in":14872,"feed_emoji":"🧮","tokens_out":13314,"duration_ms":102850,"temperature":0.7,"pith_summary":"The paper proves a foundational size statement for cohomology defined by n-step extensions: in any semi-abelian variety of algebras, the collection of all n-step extensions between any two fixed objects is small, meaning it admits a bijection to a set of representative exact sequences. This matters because extension-based cohomology objects are equivalence classes of exact sequences, and nothing in the definition guarantees that these classes form a set rather than a proper conglomerate. The proof handles one-step extensions directly by showing every short exact sequence is a retract of a fixed product shape, then extends to arbitrary length through a syzygy reduction. Variations settle the same smallness question for double extensions, crossed extensions, and Schreier extensions of monoids.","feed_headline":"All n-step extension classes are small in semi-abelian varieties","feed_subtitle":"Every n-step extension class gains a well-defined set of representatives, closing a long-standing foundational gap.","key_machinery":"The carrying mechanism is the operations-and-identities characterisation of semi-abelian varieties (Theorem 4.2): there exists $\\ell \\ge 1$, binary operations $\\alpha_1,\\dots,\\alpha_\\ell$ with $\\alpha_i(x,x)=0$, and an $(\\ell+1)$-ary operation $\\beta$ satisfying $\\beta(\\alpha_1(x,y),\\dots,\\alpha_\\ell(x,y),y)=x$. These operations make every one-step extension a retract over $Q$ of the trivial product extension $K^\\ell \\times Q \\to Q$, bounding the middle objects by subsets of $K^\\ell \\times Q$. The Syzygy Theorem (Theorem 5.1) then converts any $n$-step extension into a one-step extension of the syzygy object $\\Omega^n(Q)$, so the smallness of one-step extensions propagates to all lengths.","core_discovery":"The central claim, Theorem 5.2, is that in a category with kernels and cokernels, pullback-stable normal epimorphisms and enough normal-projective objects, smallness of $\\mathrm{Ext}^1(Q,K)$ for all $Q$ and $K$ forces smallness of $\\mathrm{Ext}^n(Q,K)$ for every $n$; because semi-abelian varieties of algebras satisfy these hypotheses, every such variety has small $n$-step extension conglomerates. The one-step case, Theorem 4.8, uses the algebraic characterisation of semi-abelian varieties to embed any middle object $X$ of a short exact sequence $0 \\to K \\to X \\to Q \\to 0$ as a subset of $K^\\ell \\times Q$ over $Q$, allowing representatives to be chosen from a single set of normal monomorphisms. This resolves a foundational gap in the classical extension-based definition of cohomology, where the collection of extensions was often simply assumed to be a set.","pith_inferences":["The proof of smallness is non-constructive: it invokes the Axiom of Choice to pick representatives of the fibre partition, so the paper does not exhibit an explicit set of extensions; a constructive version would need a choice-free way of selecting representatives.","Because smallness is established abstractly, the paper does not identify these sets with familiar invariants; for groups, the natural next step would be to show that the representative set matches the usual group-cohomology computation $\\mathrm{Ext}^n(Q,K) \\cong H^{n+1}(Q,K)$.","The monoid result suggests that the real obstruction to smallness outside semi-abelian varieties is not length but the existence of a uniform product bound for middle objects; any category with such a bound and a weakly universal normal epimorphism should admit the same argument.","If the paper's conditional criterion for Schreier extensions holds, it would yield a family of non-semi-abelian examples where higher n-step extension classes are small, extending the main theorem beyond its stated scope."],"forward_implications":["In every semi-abelian variety, the functors $\\mathrm{Ext}^n(-,K)$ from the opposite variety to $\\mathbf{Set}$ exist for each $n$, because the extension conglomerates are small.","The conglomerate of double extensions $2\\text{-}\\mathrm{Ext}(Q,K)$ is small, and consequently double central extensions form a small conglomerate without any extra commutator condition, resolving a point left open in earlier work on higher central extensions.","In strongly semi-abelian varieties, two-fold crossed extensions between any two objects form a small conglomerate.","For monoids, the one-step Schreier extension conglomerate $\\mathrm{Ext}^1_S(Q,K)$ is small for all monoids $K$ and $Q$.","If a weakly universal Schreier extension exists for every monoid $Q$, then all higher Schreier extension conglomerates are small as well."],"supporting_citations":[{"why":"Supplies the operations-and-identities characterisation of semi-abelian varieties used to construct the retraction bound for one-step extensions.","marker":"[8]"},{"why":"Provides the framework of kernels, cokernels, and normal monomorphisms and the strengthened retraction statement used in Corollary 4.6.","marker":"[35]"},{"why":"Defines semi-abelian categories, the context in which the main theorem is stated.","marker":"[26]"},{"why":"Gives the Short Five Lemma, which ensures that extension equivalence classes in semi-abelian categories are isomorphism classes.","marker":"[6]"},{"why":"Introduces the n-step extensions whose smallness is the subject of the paper.","marker":"[48]"},{"why":"Supplies the Grothendieck universe axioms that define the set-theoretic notion of smallness used throughout.","marker":"[1]"}],"fun_headline_variants":["All n-step extensions form a set in semi-abelian varieties","Semi-abelian varieties: every extension collection is a set","Theorem: In semi-abelian varieties, extensions are small","Small cohomology: semi-abelian varieties have set-sized extensions","Gap closed: n-step extensions are sets in semi-abelian varieties"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire argument depends on the characterisation of semi-abelian varieties by operations and identities, which guarantees that every one-step extension is a retract over $Q$ of $K^\\ell \\times Q$, together with the Axiom of Choice needed to select representatives from the resulting bounded collection.","fun_headline_variants_meta":{"raw":{"variants":["All n-step extensions form a set in semi-abelian varieties","Semi-abelian varieties: every extension collection is a set","Theorem: In semi-abelian varieties, extensions are small","Small cohomology: semi-abelian varieties have set-sized extensions","Gap closed: n-step extensions are sets in semi-abelian varieties"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000321,"raw_usage":{"total_tokens":1790,"prompt_tokens":913,"completion_tokens":877,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":529,"completion_tokens_details":{"reasoning_tokens":784}},"tokens_in":529,"tokens_out":877,"duration_ms":7724,"temperature":1.0,"reasoning_tokens":784,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:29:32.896686+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a semi-abelian variety $\\mathcal{V}$ and objects $K,Q$ in $\\mathcal{V}$ for which the collection of isomorphism classes of short exact sequences $0 \\to K \\to X \\to Q \\to 0$ is a proper class, so that no set of representatives exists; this would refute Theorem 4.8 and, with it, the main smallness theorem.","supporting_citations":[{"cited_title":"Bourn and G","cited_arxiv_id":null,"evidence_quote":"Supplies the operations-and-identities characterisation of semi-abelian varieties used to construct the retraction bound for one-step extensions."},{"cited_title":"Janelidze, L","cited_arxiv_id":null,"evidence_quote":"Defines semi-abelian categories, the context in which the main theorem is stated."},{"cited_title":"Bourn,Normalization equivalence, kernel equivalence and affine categories, Category The- ory, Proceedings Como 1990 (A","cited_arxiv_id":null,"evidence_quote":"Gives the Short Five Lemma, which ensures that extension equivalence classes in semi-abelian categories are isomorphism classes."},{"cited_title":"Yoneda,On Ext and exact sequences, J","cited_arxiv_id":null,"evidence_quote":"Introduces the n-step extensions whose smallness is the subject of the paper."},{"cited_title":"Artin, A","cited_arxiv_id":null,"evidence_quote":"Supplies the Grothendieck universe axioms that define the set-theoretic notion of smallness used throughout."}],"review_version":1}