{"id":"0d3b5b8a-7bf1-4623-9509-382f33829e6c","arxiv_id":"2509.02001","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper defines good endofunctors of C*-algebras and proves they give commutative monoids of generalized morphisms with bilinear composition, generalizing E-theory and KK-type constructions.","lead":"The paper creates a broad algebraic framework, based on \"good endofunctors\" of C*-algebras, for constructing generalized morphism groups, and shows it unifies K-theory, KK-theory, extension theory and E-theory as special cases. It also introduces asymptotically adjoint endofunctors, aimed at new descriptions of E-theory and K-homology.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Non-well-pointedness of A leaves its labeling, and hence the asymptotic constructions and • composition, potentially non-canonical.","rationale":"I focused on the A-labeling issue because it is a canonicity problem rather than a routine omitted diagram. The reader's weakest assumption is essentially this, though its scope is slightly overbroad: plain [A,F,B] for a C*-algebra A does not use the functor A's labeling; the dependence enters via α and κ_{K,A} in the asymptotic/E-theoretic constructions and in •. This is load-bearing because the paper's motivation is to generalize E-theory; if different good labelings of A produce different [[A,F,B]] or different •, the generalized composition is not a well-defined invariant. The concrete test with the explicit non-well-pointedness witness would either produce a counterexample to canonicity or force a direct uniqueness proof, settling the concern. I do not think the paper should be rejected: the framework is coherent and the quotient labeling is a legitimate choice; but the central E-theoretic claims cannot be accepted as fully canonical until this is resolved. Hence the reader's CONDITIONAL verdict is unchanged.","tokens_in":42309,"tokens_out":26249,"duration_ms":311071,"concrete_test":"Exhibit or exclude a second labeling. Let X=(0,∞), B=C, f(t)(x)=exp(−(tx−1)^2). Then [f]∈A C_X C is nonzero but A ev_x([f])=0 for all x, so the monicness used in Prop. 4.42 fails. Starting from the quotient labeling κ, use this kernel element to modify κ_{C_X,A} and extend by naturality to all A; check the two coherence diagrams of Def. 4.1. If κ'≠κ, compute the induced ⟨Fα⟩ on [[C,Id,C]] and • with G=A under both labelings; equality is necessary for canonicity. If no κ' exists, prove uniqueness directly without well-pointedness.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The asymptotic algebra functor A=T/T0 is not well-pointed (Remark 4.41), so Proposition 4.42 does not apply. Lemma 4.28 guarantees a unique labeling making the quotient map q:T⇒A labeled, and Remark 4.29 fixes this labeling, but it does not rule out other families {κ_{A,A}} that satisfy Definition 4.1 without making q labeled. This matters because the E-theoretic part of the paper uses A as an object of hGEFC in three places: (i) Definition 5.13 builds [[A,F,B]] using ⟨Fα⟩ with α:Id⇒A; α is labeled for the quotient labeling but need not be for an alternative one, changing the directed system; (ii) Definition 5.15/Theorem 5.16 use a counit ε:SN⇒AK into the chosen object A, so the asymptotic adjunction is sensitive to which labeled endofunctor A is; (iii) the composition • in (5.4) uses κ_{K,GK}; when G=A this is the chosen labeling of A. The paper neither proves that all good labelings of A are homotopic in hGEFC nor that these invariants are independent of the choice. Thus the advertised generalization of E-theory may be non-canonical.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a class of \"good endofunctors\" of the category of C*-algebras, i.e., decent endofunctors equipped with a compatible labeling, and organizes them into a bimonoidal category. It then defines homotopies of labeled natural transformations, constructs for every good endofunctor F and C*-algebras A, B a commutative monoid [A, FK, B] of generalized morphisms, and equips these monoids with a bilinear associative composition • that is claimed to generalize the composition in the Connes–Higson E-theory category. A further construction, the asymptotic algebra functor A = C_b([0,∞),·)/C_0([0,∞),·), is used to define asymptotic versions [[A, F, B]] and an \"asymptotic adjunction\" S ⊣_as N, yielding an isomorphism of monoids colim_n [LA, A^nK, B] ≅ colim_n [A, RA^nK, B]. The paper is largely categorical, with many explicit diagrams and a fully faithful tensor embedding; the main theorems are Propositions 5.4, 5.5, and Theorem 5.16. The presentation is self-contained relative to standard references on bimonoidal categories and C*-algebra K-theory, though some key verifications are delegated to the reader.","tokens_in":42651,"tokens_out":5986,"duration_ms":69575,"significance":"If the constructions are correct and canonical, the paper provides a substantial unifying framework: many Kasparov-type theories can be expressed as [A, F, B] for suitable good endofunctors, and the asymptotic adjunction gives a new categorical explanation of the failure of the suspension functor to have a right adjoint in the asymptotic homotopy category. The paper's strengths include its systematic categorical setup, the explicit construction of the bimonoidal category hGEFC, the fully faithful tensor embedding, and the fact that the main structural theorems are accompanied by extensive commutative diagrams. However, the advertised generalization of E-theory depends on the labeling of the asymptotic algebra functor A, and the manuscript does not establish that the resulting invariants are independent of this choice. Because this canonicality question affects the central claim, the paper needs revision before it can be accepted.","major_comments":[{"comment":"The asymptotic algebra A = T/T0 is used as an object of hGEFC in Definition 5.13, in the composition • of (5.4), and in the asymptotic adjunction of Theorem 5.16. Remark 4.41 explicitly states that A is not well-pointed, so Proposition 4.42 does not guarantee uniqueness of its labeling. Lemma 4.28 fixes one labeling by requiring the quotient map q : T ⇒ A to be labeled, but it does not rule out other labelings of A. Since [A, FK, B], [[A, F, B]], and the counit ε : SN ⇒ AK all depend on the chosen labeling, alternative labelings could produce different monoids and different adjunction isomorphisms. The paper should prove that all good labelings of A are homotopic in hGEFC, or prove that the constructions are independent of the labeling, or explicitly state that the E-theoretic claims are relative to the fixed quotient labeling.","section":"§5.2, Def. 5.13, Thm 5.16; cf. §4.5"},{"comment":"The central structural claims that [A, FK, B] is a commutative monoid and that • is associative and unital are not fully proved. In Proposition 5.4, after displaying diagrams (5.1)–(5.3), the proof says \"We leave it to the reader to deduce\" associativity, symmetry, and neutrality. Proposition 5.5 likewise leaves the unitality identities [φ]•[ι00A] = [φ] and [ι00B]•[φ] = [φ] to the reader. These are load-bearing properties of the monoid and of the composition; the derivations from the displayed perimeters should be written out or reduced to explicit lemmas (e.g., Lemma 2.11 and the rig axioms), rather than being delegated.","section":"Props. 5.4 and 5.5"},{"comment":"The proof of the asymptotic adjunction theorem uses two large diagrams to establish Ψ∘Φ = id and Φ∘Ψ = id, but the unlabeled subdiagrams are not explained, and the bookkeeping of the colimit stages (n vs. n+1, K^2 vs. K) is left implicit. In particular, the composites involving A κ_{K,A^n K}, A^{n+1}θ_B, and the compatibility with the maps A^n α must be checked carefully. The theorem is the main asymptotic result of the paper, so the proof should be expanded so that each equality follows from a named lemma or a visibly commutative diagram.","section":"Thm 5.16"}],"minor_comments":[{"comment":"Definition 4.32 defines GEFC^tt as the full subcategory generated by objects \"whose underlying endofunctors are decent\". From context this should presumably be \"tensor-type\".","section":"§4.4, Def. 4.32"},{"comment":"The uniqueness assertion in Lemma 4.28 is stated but not proved. A short argument using the fact that OA q_B is epic would make the proof complete.","section":"§4.3, Lemma 4.28"},{"comment":"Remark 4.41 says \"one can show that A := C_b(R_+)/C_0(R_+) is not well-pointed\" but gives no proof or reference. This is a key point for the labeling question and deserves a proof or a citation.","section":"§4.5, Remark 4.41"},{"comment":"The reference \"S. McLane\" should be \"S. Mac Lane\".","section":"References"},{"comment":"Several diagram labels in Section 5 are difficult to read, e.g., the composite \"F KIψ\" in the first diagram of Proposition 5.5. Please re-typeset for clarity.","section":"§5, diagrams"}],"recommendation":"major_revision","confidential_remarks":"The main concern is the non-canonical labeling of the asymptotic algebra A. This is not a disagreement with the mathematical framework but a gap in the claimed E-theory generalization; it can likely be fixed either by proving labeling-independence or by explicitly restricting to the quotient labeling and adjusting the claims. The proof gaps in Propositions 5.4 and 5.5 should also be closed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about arXiv:2509.02001. First, it is a serious categorical framework: the definition of good endofunctors, the bimonoidal category GEFC, and asymptotic adjunction are genuinely new and unify K1, KK^0, ext and E-theory. Second, the E-theory application has a gap: the asymptotic algebra functor A is not well-pointed (Remark 4.41), and the paper fixes one labeling without proving that the resulting monoids, composition, and asymptotic adjunction are independent of that choice. If there are other good labelings, the advertised generalization of E-theory may be choice-dependent.\n\nThe paper does a lot well. The construction of the monoid [A,FK,B] and the bilinear product • is coherent and carefully diagrammed. The category GEFC is complete and bimonoidal, which is a nontrivial result. The asymptotic adjunction theorem (5.16) gives a clean categorical explanation of why suspension lacks a right adjoint in hAsy. The author is upfront about the non-well-pointedness, which suggests honesty, but the consequence is not explored.\n\nThe soft spots are in proportion to how soft they actually are. The biggest is the labeling dependence. Lemma 4.28 produces a labeling via the quotient map, but uniqueness only holds for well-pointed endofunctors (Prop 4.42). For A, other labelings could exist. Since Definition 5.13 and the • composition use the chosen A as an object of hGEFC, the E-theoretic colimits and the asymptotic adjunction depend on that choice. The paper neither proves all good labelings of A are homotopic nor that the invariants are invariant. That is a load-bearing gap for the main application, though not for the general framework.\n\nSecond, the proofs of associativity and unitality in Propositions 5.4 and 5.5 are left to the reader. For the central theorem, that is a request for a referee to fill in a non-trivial diagram chase. It may be routine, but it deserves to be written out.\n\nThird, minor: the tensor-type endofunctors are adapted from known strictification results; that's fine, but the paper should credit [5,11] more explicitly (it does cite them).\n\nOverall, this is a paper for specialists in operator-algebra K-theory and categorical homotopy. If the labeling issue is resolved (or the application to E-theory is scaled back to a note about the general construction), it will be a useful reference. As it stands, I would recommend sending it to peer review, but with a requirement that the authors address the non-uniqueness of the A labeling and provide the missing coherence checks. It is not desk-rejectable: the framework is valuable and the thinking is serious.","headline":"A genuinely useful categorical framework, but the E-theory application is undercut by an unexamined choice of labeling for the asymptotic algebra functor.","tokens_in":43083,"tokens_out":4219,"would_cite":false,"duration_ms":46299,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L80","19K35","18M05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every good endofunctor of C*-algebras yields a commutative monoid of generalized morphisms with a bilinear composition generalizing E-theory.","keywords":["good endofunctors","generalized morphisms","E-theory","asymptotic homomorphisms","asymptotic adjunction","bimonoidal categories","C*-algebras","stabilization"],"falsifier":"Construct two distinct labelings of the asymptotic algebra functor A (both compatible with the quotient of C_b([0,∞),·) by C_0([0,∞),·)) for which the monoids [[A, B]] = [A, A, B] are not isomorphic for some A and B; or, on the positive side, prove directly that the quotient labeling is unique, for instance by showing that any labeling of A is forced by evaluation maps.","tokens_in":42216,"feed_emoji":"🧮","tokens_out":8027,"duration_ms":74417,"temperature":0.7,"pith_summary":"The paper sets out to build a categorical framework in which the Connes–Higson E-theory of C*-algebras is just one example of a general construction. For any 'good' endofunctor F (a decent endofunctor equipped with a compatible labeling by tensoring with C*-algebras), and any pair A, B of C*-algebras, the author forms the set [A, FK, B] of F-homotopy classes of ∗-homomorphisms and proves it is a commutative monoid. These monoids carry an associative bilinear composition that generalizes E-theory composition, and asymptotically adjoint good endofunctors induce isomorphisms of the associated colimit monoids. If correct, this gives a unified description of K-theory, KK-theory, extension theory, and E-theory within a single framework.","feed_headline":"Every good endofunctor builds a commutative morphism monoid","feed_subtitle":"The framework unifies K-theory, KK-theory, extension theory, and Connes–Higson E-theory composition.","key_machinery":"The load-bearing object is the category GEFC of good endofunctors: decent endofunctors equipped with a labeling κ_{A,F}: O_A F ⇒ F O_A that records compatibility of F with the left tensoring functor O_A = A⊗(−). The stabilization functor K (tensoring with compact operators) and the rig structure on K — with multiplication θ, unit ι00, and sum µ — supply the monoid operation + on [A, FK, B]. The asymptotic algebra functor A = C_b([0,∞),·)/C_0([0,∞),·) is the motivating example; it is a good endofunctor obtained by quotient, although not well-pointed. Asymptotic adjunctions generalize the relationship between suspension and the Roe-algebra endofunctor, and Theorem 5.16 turns an asymptotic adju","core_discovery":"The central claim is that the class of good endofunctors of C*-algebras forms a tight bimonoidal category hGEFC, and that every object F determines a commutative monoid [A, FK, B] of generalized morphisms together with a bilinear associative composition •. The definition is built so that the asymptotic algebra functor A reproduces the Connes–Higson homotopy category of asymptotic homomorphisms, and composition in the E-theory category appears as a special case. The paper further introduces asymptotic adjunctions S ⊣_as N between good endofunctors, defined by a unit η : Id ⇒ NS and a counit ε : SN ⇒ AK, and proves (Theorem 5.16) that any such adjunction gives mutually inverse monoid isomorphi","pith_inferences":["If alternative labelings of non-well-pointed functors like A exist, the framework would describe a family of E-theory-like categories rather than a single canonical one; this dependence is flagged but not resolved in the paper.","The bracketing/EV machinery suggests that any construction in the bimonoidal category of C*-algebras can be transferred to good endofunctors, so one could test whether replacing the compact-operator rig K by another rig in hC* yields twisted generalized morphism monoids.","The asymptotic adjunction formalism may produce a right asymptotic adjoint for suspension, circumventing the known nonexistence of an ordinary right adjoint in the asymptotic homotopy category."],"forward_implications":["Taking F to be the identity, suspension, double, corona, or asymptotic algebra functor recovers K_1, KK_0, extension groups, and Connes–Higson E-theory as the monoids [A, FK, B].","The bilinear composition • provides a common generalization of the Kasparov product and the composition of asymptotic homomorphisms.","Any asymptotic adjunction S ⊣_as N yields an isomorphism of monoids colim_n [SA, A^nK, B] ≅ colim_n [A, NA^nK, B], giving a KK-like model for E-theory.","Endofunctors admitting inversion — including the suspension functor — produce abelian groups rather than just monoids at the level of [A, F, B]."],"supporting_citations":[{"why":"Defines the asymptotic algebra functor and the Connes–Higson E-theory that the paper generalizes.","marker":"[6]"},{"why":"Provides the rig structure on the compact operators and homotopy lemmas underlying the monoid operation.","marker":"[8]"},{"why":"Supplies the definitions and Laplaza axioms for bimonoidal categories used throughout Sections 1–4.","marker":"[10]"},{"why":"Prior result identifying E_0(SA,B) with generalized morphisms of the Roe-algebra endofunctor, motivating this framework.","marker":"[13]"},{"why":"Prior description of E-theory composition and preliminary axioms for the endofunctors.","marker":"[14]"},{"why":"KK-like picture for E-theory which the asymptotic adjunction generalizes.","marker":"[15]"},{"why":"Shows the suspension functor has no right adjoint in the asymptotic homotopy category, motivating asymptotic adjunctions.","marker":"[19]"},{"why":"Provides the double and corona endofunctors used as examples of generalized morphisms.","marker":"[3]"}],"fun_headline_variants":["Good endofunctors yield commutative morphism monoids","C*-algebra endofunctors: a unified morphism framework","Generalized morphisms from good endofunctors","New monoids generalize E-theory composition","Asymptotic adjunctions unify E-theory and K-homology"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The construction depends on the quotient labeling of the asymptotic algebra functor A being canonical; since A is not well-pointed, the paper does not prove uniqueness of this labeling, so the monoids and adjunctions built from it could in principle depend on the choice.","fun_headline_variants_meta":{"raw":{"variants":["Good endofunctors yield commutative morphism monoids","C*-algebra endofunctors: a unified morphism framework","Generalized morphisms from good endofunctors","New monoids generalize E-theory composition","Asymptotic adjunctions unify E-theory and K-homology"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000146,"raw_usage":{"total_tokens":986,"prompt_tokens":681,"completion_tokens":305,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":425,"completion_tokens_details":{"reasoning_tokens":222}},"tokens_in":425,"tokens_out":305,"duration_ms":3631,"temperature":1.0,"reasoning_tokens":222,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T11:58:56.418384+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct two distinct labelings of the asymptotic algebra functor A (both compatible with the quotient of C_b([0,∞),·) by C_0([0,∞),·)) for which the monoids [[A, B]] = [A, A, B] are not isomorphic for some A and B; or, on the positive side, prove directly that the quotient labeling is unique, for instance by showing that any labeling of A is forced by evaluation maps.","supporting_citations":[{"cited_title":"Johnson, D","cited_arxiv_id":null,"evidence_quote":"Supplies the definitions and Laplaza axioms for bimonoidal categories used throughout Sections 1–4."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior result identifying E_0(SA,B) with generalized morphisms of the Roe-algebra endofunctor, motivating this framework."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior description of E-theory composition and preliminary axioms for the endofunctors."}],"review_version":1}