{"id":"760aab19-56fc-4293-8762-cb411d293e08","arxiv_id":"2501.12510","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For each relative pseudomonad T, the paper builds a terminal bicategory of T-pseudoalgebras, embeds the Kleisli bicategory into it, and characterizes pseudoalgebras for free cocompletion pseudomonads as cocomplete categories.","lead":"The paper introduces pseudoalgebras for relative pseudomonads, the two-dimensional analogue of algebras for relative monads, and proves the bicategory of such pseudoalgebras is the terminal resolution of the pseudomonad. It uses this to prove that the bicategory of distributors is biequivalent to the 2-category of presheaf categories, a folklore result previously lacking a rigorous proof.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 7.8's uniqueness argument is the load-bearing step for the cocompleteness characterization; its intricate diagram chase is not independently verified.","rationale":"The reader's weakest_assumption correctly identifies Theorem 7.14 and its dependence on Lemma 7.8 as the least secure part of the paper. My stress-test sharpens this into a concrete internal concern: the uniqueness proof in Lemma 7.8 involves a non-trivial diagram chase whose key step is not fully detailed. I do not claim the step is wrong; the paper's overall development is coherent, and the authors are careful elsewhere. However, the flagship distributors-presheaf biequivalence depends on this lemma, and the authors themselves flag the technique as fragile (Remark 7.20). Since no demonstrated error exists, and since machine-checked proofs are absent, the honest recommendation is to keep the ACCEPT verdict while noting that a formal or independent verification of Lemma 7.8 would materially strengthen the paper. I therefore do not propose changing the verdict, only adding this as the single load-bearing concern to settle.","tokens_in":55790,"tokens_out":21917,"duration_ms":219279,"concrete_test":"Formalize the proof of Lemma 7.8 in a proof assistant with an existing bicategory or category theory library (e.g., UniMath or Lean's mathlib), encoding relative pseudomonads, pseudoalgebras, and the (-)^⊤ construction. Machine-check the uniqueness of the mediating cocone morphism, in particular the evaluation of diagrams (31) and (32) at the terminal object t. If the formalization cannot close the uniqueness step, the lemma needs repair. A lighter check: instantiate the argument for the free initial-object completion (Φ = {∅}) on a finite category A, explicitly compute the colimit constructed by Lemma 7.8, and verify that the mediating morphism κ agrees with the unique morphism forced by the universal property of the initial object.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The flagship application (the biequivalence between distributors and the 2-category of presheaf categories, Corollary 6.15 + Theorem 7.14) rests on Theorem 7.14, which in turn rests on Lemma 7.8. Lemma 7.8 constructs colimits in an arbitrary pseudoalgebra for a relative pseudomonad on CAT, and its proof of uniqueness of morphisms of cocones is intricate: it evaluates diagrams (31) and (32) at the terminal object t of T(D) and concludes that a certain morphism fa(·) is the identity by the universal property of t. This step mixes the non-strict pseudoalgebra coherence cells (ˆa, ˜a) with the pseudonaturality of the extension operator, and the text does not spell out how the pseudonaturality 2-cells compose along the evaluation. The authors themselves flag the fragility of the technique in Remark 7.20, noting it does not extend to enriched categories and may even fail there. If this uniqueness step contains a hidden coherence error, then Theorem 7.14 and the distributors-presheaf biequivalence are unsupported, even though the general terminality result (Theorem 6.10) and the Kleisli-to-pseudoalgebra embedding might still hold. This is a concern about internal correctness, not about scope: the paper's own acknowledged limitation marks Lemma 7.8 as the least secure pillar of the central application.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a theory of pseudoalgebras for relative pseudomonads. It defines bicategories of pseudoalgebras with strict, pseudo, lax, and colax morphisms; constructs a free--forgetful relative pseudoadjunction; proves doctrinal adjunction and transport of structure; extends lax-idempotence to the relative setting; and establishes universal properties for the Kleisli and pseudoalgebra resolutions of a relative pseudomonad, with the pseudoalgebra resolution shown to be 2-terminal. The main applications are a characterisation of pseudoalgebras for free cocompletion relative pseudomonads as categories with the appropriate class of colimits (Theorem 7.14), yielding a proof that the bicategory of distributors is biequivalent to the 2-category of presheaf categories, and a correspondence between cocompletion-relative monads and cocontinuous monads on free cocompletions (Theorem 8.3). The paper is carefully written, with extensive diagrammatic proofs and explicit acknowledgement of its own open points.","tokens_in":56065,"tokens_out":9782,"duration_ms":100589,"significance":"If the central results hold, the paper fills a genuine gap in two-dimensional monad theory: relative pseudomonads had no systematic algebra theory, and the paper supplies one, including the expected universal properties and the first rigorous proof, to the authors' knowledge, of the folklore biequivalence between distributors and presheaf categories. The treatment of lax-idempotence is notably more subtle than in the non-relative case, and the authors are honest about the limits of their techniques, for example in Remark 5.24 and Remark 7.20. The paper does not rely on machine-checked proofs, but its diagrammatic arguments are detailed and the main structure is coherent. The principal reservation is that the characterisation in Theorem 7.14 depends on Lemma 7.8, a long coherence argument whose key uniqueness step is not fully supported as written.","major_comments":[{"comment":"The uniqueness argument in Lemma 7.8 is load-bearing for Theorem 7.14, and as written it is not fully supported. The proof evaluates diagram (31) at the terminal object t of T(D) and then pastes on diagram (32); however, in (32) the left vertical arrow is evaluated at i_{D^⊤}(⊤), an object of T(D^⊤), while t is an object of T(D). It is not immediately evident that these objects can be identified, and the text gives no explicit justification for doing so. If they are not identified, the clockwise composite whose equality with (˜a^{−1}_f)^a(t) is claimed is not even well-formed. Please either repair this step or expand the diagram chase so that the identification of the two objects and the composition of the pseudonaturality 2-cells along the evaluation at t is spelled out. Because Lemma 7.8 is the only route by which the paper obtains colimits in arbitrary pseudoalgebras in Theorem 7.14, this is not a merely stylistic issue.","section":"§7.1, Lemma 7.8 (proof, diagrams (31)–(32))"},{"comment":"The proof of Theorem 7.14 chains Lemma 7.8 through Proposition 7.12 and Corollary 7.13 into Corollary 5.8 to prove that the comparison pseudofunctor is locally an equivalence. This is structurally sound, but it means that local full faithfulness of Φ-COC → PsAlg(Φ) is obtained only after Lemma 7.8 has been fixed. Please make this dependency explicit in the proof, and state explicitly that Corollary 7.13 uses Proposition 7.12, and hence Lemma 7.8, at the point where it asserts that fa is the left extension of f. A reader who is not prepared to accept Lemma 7.8 as a black box currently has no way to isolate the rest of the argument.","section":"§7.2, Theorem 7.14 (dependency chain)"}],"minor_comments":[{"comment":"The notation J(D^⊤) = D^⊤ = (JD)^⊤ is used without comment. Since J is injective on objects, the identification is harmless, but it should be stated explicitly before the diagram chase begins.","section":"§7.1, proof of Lemma 7.8"},{"comment":"The labels 'A unit.' and 'A unit.′' appear in the diagrams without definitions. Please give the precise displayed axiom being invoked, for instance part (6) of Definition 3.1 or Lemma 3.15, so that the reader can check the two labels separately.","section":"§7.1, proof of Lemma 7.8"},{"comment":"Since Theorem 7.14 is the flagship application and Remark 7.20 explicitly limits the technique to ordinary categories, the statement of Theorem 7.14 should carry a cross-reference to Remark 7.20 and should state clearly that the enriched case is outside the scope of the proof.","section":"§7.2, Theorem 7.14 and Remark 7.20"},{"comment":"The definition of Res(T) uses strict equalities MLX = L′X and R = R′M. This is internally consistent, but the words 'bi-initial' and '2-terminal' may suggest a weaker bicategorical universal property. Remark 6.8 is helpful; consider adding a caution immediately after Definition 6.1 that the universal properties are relative to this strict class of resolution morphisms.","section":"§6.1, Definition 6.1 and Remark 6.8"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely correct and valuable, and my recommendation is driven by the proof of Lemma 7.8 rather than by doubts about the overall architecture. The authors should be encouraged either to produce a machine-checked version of the diagram chases in Section 7.1 or to provide a substantially expanded hand-checked derivation of the uniqueness step. If Lemma 7.8 is repaired or replaced by a more robust argument, the paper should be acceptable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a serious and significant paper. It supplies the missing theory of pseudoalgebras for relative pseudomonads, and the central universal properties—bi-initiality of the Kleisli resolution and 2-terminality of the pseudoalgebra resolution—are new, well-motivated, and, as far as I can check, correctly proven. Theorem 6.10 is particularly solid: the proof uses local faithfulness of the forgetful pseudofunctor plus the canonical pseudoalgebra structure on right pseudoadjoints, and it does not borrow from the later cocompleteness material. Corollary 6.11, the fully faithful embedding of the Kleisli bicategory into PsAlg(T), is a nice coherence result that stands alone.\n\nThe paper gives credit where it is due: the comparison with Marmolejo–Wood no-iteration pseudoalgebras (Section 3.2) is careful, and the authors are explicit about what is not known (Remark 5.24, Remark 7.20). The citation pattern is normal for the subject; the heavy reliance on FGHW18 is justified because that paper is the source of the relative pseudomonad formalism.\n\nThe soft spot is exactly where the stress-test places it: Lemma 7.8. The uniqueness part of the colimit construction evaluates the large diagrams (31) and (32) at the terminal object of T(D), and concludes that a composite morphism is the identity from the universal property of that terminal object. The chase is intricate and mixes the pseudoalgebra coherence cells with pseudonaturality of the extension operator; the text does not fully spell out how the pseudonaturality 2-cells compose along the evaluation. I did not find a definite error, but I also could not verify every step from the text alone. Consequently, Theorem 7.14—the characterization of pseudoalgebras for free cocompletions as the Φ-cocomplete categories—is the least secure pillar. If Lemma 7.8 has a hidden coherence bug, the distributors–presheaf biequivalence would lose its proof, though the general terminality theorem and the Kleisli embedding would survive.\n\nThe Section 8 application (relative monads as cocontinuous monads) is a nice payoff but inherits the dependence on Theorem 7.14. For a reader, the value is mainly in Sections 3–6, which lay out a solid framework; Section 7 should be read with extra care.\n\nThis paper deserves a serious referee. The right referee will spend most of their time on Lemma 7.8, ideally redoing the uniqueness chase or at least testing it on the free cocompletion example. I would engage with it and recommend the same.","headline":"Important and mostly sound; the cocompleteness characterization rests on a delicate coherence chase that needs careful checking.","tokens_in":56549,"tokens_out":3572,"would_cite":true,"duration_ms":35730,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18C15","18C20","18D60","18M50","18N10","18N15","18N20"],"pacs":[],"model":"deepseek-v4-flash","headline":"For every relative pseudomonad, its pseudoalgebras form a terminal resolution, and the Kleisli bicategory embeds into them.","keywords":["relative pseudomonad","pseudoalgebra","Kleisli bicategory","presheaf construction","cocomplete category","distributor","coherence theorem","lax-idempotent pseudomonad"],"falsifier":"Attempt to construct a P-pseudoalgebra structure on a locally small category that lacks some small colimit; the pseudoalgebra axioms require a left extension of every functor from a small category, and if such an extension can be defined despite the missing colimit, Theorem 7.14 is false. Conversely, the theorem's proof constructs the missing colimit from the extension operator, so checking this single example settles the matter.","tokens_in":55615,"feed_emoji":"🧩","tokens_out":12619,"duration_ms":100876,"temperature":0.7,"pith_summary":"The paper builds a two-dimensional algebra theory for relative pseudomonads, which are monad-like structures that need not be endofunctors, the motivating example being the presheaf construction on small categories. It defines pseudoalgebras for such a T and constructs the free–forgetful relative pseudoadjunction from the bicategory PsAlg(T) of T-pseudoalgebras, proving it is 2-terminal among all resolutions of T: every relative pseudoadjunction inducing exactly T factors through it, with the mediating pseudofunctor unique up to isomorphism. Consequently, the Kleisli bicategory embeds fully faithfully into PsAlg(T) as the free pseudoalgebras, and when the codomain is a 2-category this yields a coherence theorem: Kl(T) is biequivalent to the full sub-2-category of free pseudoalgebras. For the presheaf relative pseudomonad, the paper proves that pseudoalgebras are exactly locally small categories with a choice of small colimits, so the bicategory of distributors is biequivalent to the full sub-2-category of presheaf categories inside the 2-category of cocomplete categories. Along the way it extends doctrinal adjunction, transport of structure, and lax-idempotence to the relative setting.","feed_headline":"Relative pseudomonads gain canonical bicategories of algebras","feed_subtitle":"For the presheaf construction, pseudoalgebras are exactly cocomplete categories","key_machinery":"The central object is a T-pseudoalgebra for a J-relative pseudomonad T: an object A of the codomain together with an extension operator (−)^a : E[JX,A] → E[TX,A] and invertible 2-cells â and ã satisfying associativity and unit laws (Definition 3.1). The load-bearing mechanism is the resolution 2-category Res(T), whose objects are relative pseudoadjunctions inducing exactly T, together with the two universal constructions: the Kleisli bicategory (bi-initial, Theorem 6.3) and the pseudoalgebra bicategory (2-terminal, Theorem 6.10). The comparison pseudofunctor IT : Kl(T) → PsAlg(T) of Corollary 6.11 is what converts the coherence statement: it is fully faithful and, when E is a 2-category, exhibits Kl(T) as biequivalent to the full sub-2-category of free pseudoalgebras (Corollary 6.15). For the presheaf application, the pivotal identity is that a functor D → A with small domain D has a colimit exactly when it extends along the free cocone inclusion D ↪ D^⊤, and Lemma 7.8 uses the pseudoalgebra structure to construct that extension, thereby building colimits in any pseudoalgebra.","core_discovery":"Theorem 6.10 states that for any relative pseudomonad T, the free–forgetful relative pseudoadjunction from the bicategory PsAlg(T) of T-pseudoalgebras and pseudomorphisms is 2-terminal in the 2-category Res(T) of resolutions of T: every relative pseudoadjunction inducing exactly T has a unique morphism to the pseudoalgebra resolution, and the only 2-cell from that morphism to itself is the identity. Consequently (Corollary 6.11) there is a unique fully faithful pseudofunctor Kl(T) → PsAlg(T) whose image is the free pseudoalgebras, and when the codomain is a 2-category this exhibits Kl(T) as biequivalent to the full sub-2-category of free pseudoalgebras (Corollary 6.15). In the specific case of the presheaf relative pseudomonad P, the paper proves that every P-pseudoalgebra is precisely a locally small category equipped with chosen small colimits (Theorem 7.14), so the bicategory Dist of distributors is biequivalent to the full sub-2-category of the 2-category COC spanned by presheaf categories. The proof of Theorem 7.14 constructs colimits inside any pseudoalgebra using the extension operator and the one-point-cocone category D^⊤, and it relies on Lemma 7.8, which assumes D^⊤ lies in the domain and that the unit preserves terminal objects; as Remark 7.20 notes, the technique is not known to extend to enriched categories.","pith_inferences":["Beyond the paper's claims: the 2-terminality theorem gives a uniform strictification recipe, so any Kleisli bicategory for a relative pseudomonad valued in a 2-category is automatically biequivalent to the concrete 2-category of free algebras; other size-sensitive constructions (free coproducts, Ind-completions, sifted colimits) should admit the same coherence theorem without case-by-case work.","The limitation flagged in Remark 7.20 suggests an enriched analogue will need a different technique than the D^⊤-cocone construction; if developed, it would likely give enriched coherence theorems for enriched distributors and cocomplete enriched categories.","One testable extension: the stronger notion of lax-idempotent relative pseudomonad (all pseudoalgebras lax-idempotent) may be the right one for future work; if it implies every colax morphism between pseudoalgebras is pseudo, it would fill the gap noted in Remark 5.24.","The correspondence of Theorem 8.3 between relative monads and cocontinuous monads is likely an instance of a more general relative monadicity that could be checked by verifying the triangle comparison for Ind-completions and sifted colimits directly."],"forward_implications":["For every relative pseudomonad T, the bicategory PsAlg(T) of T-pseudoalgebras is the terminal resolution of T: every relative pseudoadjunction inducing exactly T factors through it, and the mediating pseudofunctor is unique up to unique isomorphism (Theorem 6.10).","The Kleisli bicategory Kl(T) embeds fully faithfully into PsAlg(T) as the free pseudoalgebras (Corollary 6.11); when the codomain is a 2-category, Kl(T) is biequivalent to the full sub-2-category of free pseudoalgebras (Corollary 6.15).","For the presheaf relative pseudomonad P, the pseudoalgebras are precisely locally small categories equipped with chosen small colimits, so the bicategory Dist of distributors is biequivalent to the full sub-2-category of presheaf categories in the 2-category of cocomplete categories (Theorem 7.14 and Corollary 6.15).","For any class Φ of small categories closed under adjoining a terminal object, the free Φ-cocompletion relative pseudomonad has exactly the Φ-cocomplete categories as pseudoalgebras, and every pseudoalgebra is lax-idempotent (Theorem 7.14, Corollary 7.13).","Monads relative to the free Φ-cocompletion of a small category A are equivalent to Φ-cocontinuous monads on Φ(A), and the equivalence commutes with the taking of algebras (Theorem 8.3)."],"supporting_citations":[{"why":"Supplies the definition of relative pseudomonads and Kleisli bicategories, which the paper extends with pseudoalgebras.","marker":"[FGHW18]"},{"why":"Establishes the one-dimensional theory of relative monad algebras and their resolutions, whose universal property the paper lifts to two dimensions.","marker":"[ACU15]"},{"why":"Develops the formal theory of relative monads and strict algebras, providing the comparison template for Corollary 6.11.","marker":"[AM24]"},{"why":"Introduces no-iteration pseudomonads and their pseudoalgebras, which the paper proves equivalent to its own pseudoalgebras for identity-relative pseudomonads.","marker":"[MW13]"},{"why":"Originates doctrinal adjunction, which Theorem 4.1 extends to relative pseudomonads.","marker":"[Kel74]"},{"why":"Defines the bicategory of distributors, which the paper identifies as the Kleisli bicategory of the presheaf construction.","marker":"[Bén73]"},{"why":"Provides the free cocompletion and left extension machinery used to characterise Φ-cocompletion pseudoalgebras in Section 7.","marker":"[Kel82]"}],"fun_headline_variants":["Pseudoalgebras for relative pseudomonads are terminal resolutions","Relative pseudomonads get canonical free-forgetful adjunctions","Presheaf pseudoalgebras are exactly cocomplete categories","Terminal resolutions: pseudoalgebras for relative pseudomonads","Coherence: distributors biequivalent to presheaf categories"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The identification of free-cocompletion pseudoalgebras with cocomplete categories (Theorem 7.14) rests on the assumption that every relevant diagram shape D and its one-point-cone extension D^⊤ belong to the domain of the relative pseudomonad, and that the unit preserves terminal objects; this holds for ordinary categories but the authors state they do not know how to extend it to enriched categories.","fun_headline_variants_meta":{"raw":{"variants":["Pseudoalgebras for relative pseudomonads are terminal resolutions","Relative pseudomonads get canonical free-forgetful adjunctions","Presheaf pseudoalgebras are exactly cocomplete categories","Terminal resolutions: pseudoalgebras for relative pseudomonads","Coherence: distributors biequivalent to presheaf categories"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000236,"raw_usage":{"total_tokens":1558,"prompt_tokens":1056,"completion_tokens":502,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":672,"completion_tokens_details":{"reasoning_tokens":416}},"tokens_in":672,"tokens_out":502,"duration_ms":4313,"temperature":1.0,"reasoning_tokens":416,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T17:06:51.264695+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Attempt to construct a P-pseudoalgebra structure on a locally small category that lacks some small colimit; the pseudoalgebra axioms require a left extension of every functor from a small category, and if such an extension can be defined despite the missing colimit, Theorem 7.14 is false. Conversely, the theorem's proof constructs the missing colimit from the extension operator, so checking this single example settles the matter.","supporting_citations":[],"review_version":1}