{"id":"d32dea4f-8de6-471a-9aa3-b8404e0d1a9b","arxiv_id":"2507.04138","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Every Borel-overt fiberwise quasi-Polish groupoid is Borel equivalent to an open Polish groupoid, hence to an action groupoid of a Polish group action; Borel-overt classwise Polish equivalence relations are Borel bireducible with orbit equivalence relations of free Polish group actions.","lead":"This paper shows that Borel equivalence relations and groupoids carrying a uniformly Borel family of Polish (or quasi-Polish) topologies on each class or component can always be represented, up to Borel equivalence, by the orbit structures of actual Polish group actions. It turns the orbitwise topology behind the Becker-Kechris theorem into an abstract axiom system, proves the axioms are complete, and derives Effros, Vaught, and open mapping results for the new framework.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 4.3.1's factorization step is underjustified: the displayed μ×μ-quantifier identity needs a nontrivial calculation with the simplicial fiberwise topology, and C'_ij is never defined.","rationale":"The paper is a careful and ambitious contribution, and the reader's conditional verdict is appropriate. The most load-bearing point is indeed Lemma 4.3.1, as the reader identified. However, I do not regard the uniform/non-uniform distinction itself as the core defect: the lemma is explicitly designed to bridge that gap. The real stress point is the internal proof of the lemma, where a displayed equality of fiberwise category quantifiers is asserted with insufficient justification and an auxiliary set C'_ij appears without definition. This is a concrete place where the argument could fail even if the uniform/non-uniform question is settled, and it is directly needed for Theorems 4.3.2 and 4.4.10. I do not claim the lemma is false, only that the current proof is incomplete at this step. The final dependence on the unpublished [Che19, 1.2] for the action-groupoid corollaries is a legitimate secondary concern, but it does not affect the internal proof of Theorem 4.4.10. A fully expanded proof of Lemma 4.3.1, or an independent derivation via Corollary 4.3.4, would resolve the main worry.","tokens_in":57632,"tokens_out":20787,"duration_ms":221866,"concrete_test":"Write out Lemma 4.3.1 in full detail: define C'_ij explicitly and prove, for each basic U_i, the identity μ^{-1}(U_i) = ⋃_j ((V_ij * A'_ij) ×_Y (B'_ij * C'_ij)) using the canonical ∂s-fiberwise topology on G⁴ from Remark 3.5.21. In particular, verify that the fiber of μ×μ : G⁴ → G×_X G over a composable pair is nonmeager in itself, and that the replacement of C_ij by a τ-fiberwise open C'_ij does not change the quantifier. If this computation reveals an additional matching obstruction, then Lemma 4.3.1 needs a different proof and Theorem 4.4.10 is not established as written.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Theorem 4.4.10 depends on Lemma 4.3.1 to convert an arbitrary Borel-overt fiberwise quasi-Polish groupoid into a uniformly componentwise quasi-Polish one on a comeager full subgroupoid. The proof of Lemma 4.3.1 contains a step that is not fully justified. Starting from μ4^{-1}(U_i) = ⋃_j (V_ij ×_X A_ij ×_X B_ij ×_X C_ij), the proof replaces A_ij, B_ij by Borel fiberwise open approximations A'_ij, B'_ij, and then asserts μ^{-1}(U_i) = ⋃_j ((V_ij * A'_ij) ×_Y (B'_ij * C'_ij)). Here C'_ij is introduced with no definition, and more importantly the equality requires the ∃*_{μ×μ}-quantifier over the fiber of the quaternary multiplication map to factor as a product of the separate Vaught transforms V*A' and B'*C'. This is not automatic: the fiber over a composable pair (p,q) involves an extra matching condition between the middle objects of the two factorizations, and its nonmeagerness is precisely the kind of hom-set/Π⁰₂ issue that Remark 3.3.2 leaves open. Without an explicit calculation using the canonical simplicial fiberwise topologies of Remark 3.5.21 (rather than the product of the σ- and τ-fiberwise topologies), Lemma 4.3.1 is underproved, and the central realization theorem 4.4.10 lacks a complete route from fiberwise data to uniform difference continuity.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops an axiomatic theory of Borel groupoids equipped with componentwise (quasi-)Polish topologies, in both fiberwise and classwise forms. The central results are: (1) a topological realization theorem (Theorem 4.1.5 / 1.3.3) showing that Borel-overt uniformly componentwise quasi-Polish groupoids admit compatible global open quasi-Polish groupoid topologies; (2) a comeager-subgroupoid theorem (Lemma 4.3.1, Theorem 4.3.2) intended to reduce arbitrary Borel-overt fiberwise quasi-Polish groupoids to uniformly componentwise ones; (3) a Polishability criterion (Theorem 4.4.5) and a passage to open Polish groupoids (Theorem 4.4.10); and (4) corollaries asserting that every Borel-overt fiberwise quasi-Polish groupoid is Borel equivalent to an action groupoid of a Polish group action, and that every Borel-overt classwise quasi-Polish equivalence relation is Borel bireducible with a free Polish group action. Along the way the paper generalizes Vaught transforms, Effros's theorem, the open mapping theorem, and related tools to the quasi-Polish groupoid setting.","tokens_in":57769,"tokens_out":15307,"duration_ms":163539,"significance":"If the proofs are correct, this is a substantial and well-motivated contribution to descriptive set theory and topological groupoid theory. The paper proposes a clear axiomatic framework for 'componentwise Polish structure', proves a genuine converse to the Becker--Kechris direction for the overt class, and shows that the resulting machinery can be used to derive several classical theorems in a unified way. The use of quasi-Polish spaces, sigma-topologies, and point-free-style algebraic manipulation is innovative, and the paper is honest about its limitations, explicitly stating open questions (Section 1.6) and providing counterexamples showing the necessity of the overtness and uniform-continuity assumptions (Examples 4.1.9 and 4.4.14). The proofs are generally detailed and the architecture of the argument is coherent. The main reservations concern one underjustified step in the proof of the key lemma (Lemma 4.3.1) and the dependence of the headline action-groupoid conclusion on an unpublished same-author preprint.","major_comments":[{"comment":"The displayed computation after 'thus by Pettis's theorem' is the load-bearing step of the proof and is underjustified. The symbol C'_ij appears in the conclusion but is never defined; presumably it is Cij or a Borel approximation to it. More substantively, the equality μ^{-1}(U_i) = ∃∗_{μ×μ}(μ4^{-1}(U_i)) = ⋃_j ((V_ij ∗ A'_ij) ×_Y (B'_ij ∗ C'_ij)) requires a proof that the ∃∗_{μ×μ}-quantifier over the fiber of the quaternary multiplication factors as a product of the two Vaught transforms V∗A' and B'∗C'. The fiber of μ×μ over a composable pair (p,q) is the product of the fibers of μ over p and μ over q, but the nonmeagerness of a subset of that fiber is not equivalent to nonmeagerness of its projections without an explicit Kuratowski–Ulam argument using the canonical simplicial fiberwise topologies of Remark 3.5.21. The current proof cites only Pettis's theorem and the separate fiberwise Baire property, which do not by themselves justify the factorization. Since Lemma 4.3.1 is the only step that upgrades an arbitrary Borel-overt fiberwise quasi-Polish groupoid to one with uniformly fiberwise continuous differences, Theorem 4.4.10 and Corollaries 4.4.11–4.4.12 rest on this missing calculation. Please supply the calculation or restructure the proof.","section":"§4.3, proof of Lemma 4.3.1"},{"comment":"The advertised conclusion that every Borel-overt fiberwise quasi-Polish groupoid is Borel equivalent to an action groupoid of a Polish group action is obtained by composing Theorem 4.4.10 with [Che19, 1.2], which is an unpublished same-author preprint (arXiv:1908.03268). The paper gives no information about the publication status of [Che19] or a proof of the cited statement. This dependency is load-bearing for the headline result, not an incidental remark. Please either include a proof of [Che19, 1.2] (or a reference to a published version), or clearly state in the introduction and abstract that the action-groupoid conclusion is conditional on that preprint.","section":"§4.4, Corollary 4.4.11 (also §1.2)"}],"minor_comments":[{"comment":"In the same proof, the set C'_ij should be defined or replaced by C_ij; as written, it is an undefined symbol in a central display.","section":"§4.3, proof of Lemma 4.3.1"},{"comment":"The phrase 'free Borel action of a Polish group G⟳Y' should be 'free Borel action of a Polish group' when the group is named G; the current wording is slightly ambiguous about whether G acts freely or the action is merely Borel.","section":"§1.1, Theorem 1.1.2"},{"comment":"The notation N for both the discrete natural numbers and the one-point compactification N∪{∞} is confusing; a different symbol for the compactification would improve readability.","section":"§4.1, Example 4.1.9"},{"comment":"The discussion of the open questions about the uniform versus non-uniform versions and about Polishability of classwise Polish equivalence relations is welcome and well placed; it would be helpful to state explicitly in Section 1.3 that Theorem 1.3.3 is not claimed without the uniformity assumption, even though this is implicit in the definitions.","section":"§1.6"}],"recommendation":"major_revision","confidential_remarks":"The main technical concern is the incomplete justification in the proof of Lemma 4.3.1. If the author can supply the missing Kuratowski–Ulam/simplicial-fiberwise calculation, the paper is likely acceptable; without it, the central realization theorem lacks a complete proof. The dependence on the unpublished [Che19] for the headline action-groupoid statement should also be resolved before acceptance. These are both fixable within the manuscript's scope, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read through the whole thing. It's a substantial paper—the framework of Borel-overt fiberwise quasi-Polish groupoids and the componentwise sigma-topologies is real, and the structural results (Theorems 4.1.3, 4.1.5, plus the Effros and open-mapping generalizations in 4.5) are genuinely new and largely carefully proved. The proof architecture is coherent, and the paper is honest about its open ends (Section 1.6 and Remark 4.1.10). If the main theorem holds, it's a nice converse to Becker–Kechris.\n\nThe soft spot is exactly where the reader and the stress-test note put it: Lemma 4.3.1. As written, the step asserting μ^{-1}(U_i) = ⋃_j ((V_ij * A'_ij) ×_Y (B'_ij * C'_ij)) is not justified. There's an apparent typo—C'_ij appears without definition—and more substantively, the ∃*_{μ×μ} quantifier over the quaternary multiplication fiber doesn't factor into two independent Vaught transforms. For a given composable pair (k,l), a factorization k=va and l=bc must satisfy the middle matching σ(a)=τ(b). The separate nonmeagerness of the fibers over k and l doesn't imply that the matching subset is nonmeager over (k,l). That's a genuine hom-set/Π⁰₂ issue, of the sort the paper itself says is open in Remark 3.3.2. The proof would need a calculation using the simplicial fiberwise topologies of Remark 3.5.21 (or something equivalent) to control the middle objects, and it isn't there. So Theorem 4.4.10 currently rests on an unproved lemma.\n\nThis is not a fatal objection if the gap is repairable—my guess is it is, since the surrounding machinery is well-built—but it is load-bearing. The final corollary also leans on the unpublished [Che19], which is a secondary concern, not a fatal one.\n\nWho should read it? Specialists in descriptive set theory working on Polish groupoids and the Becker–Kechris correspondence will want to see it; the axiomatic definitions alone are worth having. I'd send it to a referee, not desk reject, but with a pointed request to scrutinize Lemma 4.3.1. As it stands, the main theorem is conditional.","headline":"A serious attempt at a Becker–Kechris converse that is worth refereeing, but Lemma 4.3.1 has a genuine gap in the factorization step.","tokens_in":58614,"tokens_out":8905,"would_cite":false,"duration_ms":86345,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["03E15","22A22","22F10"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every Borel-overt fiberwise quasi-Polish groupoid is Borel equivalent to a Polish group action.","keywords":["Polish groupoid","quasi-Polish space","Borel equivalence relation","Borel groupoid","componentwise topology","Becker–Kechris theorem","Vaught transform","topological realization"],"falsifier":"Find a Borel-overt fiberwise quasi-Polish groupoid for which no componentwise comeager full subgroupoid has uniformly fiberwise continuous differences, or find a Borel-overt fiberwise Polish groupoid whose connectedness relation is not Borel bireducible with the orbit equivalence relation of any free Polish group action. The paper explicitly leaves open whether uniform and non-uniform continuity of differences are equivalent, so the first is the most direct test.","tokens_in":57183,"feed_emoji":"🔄","tokens_out":8821,"duration_ms":91722,"temperature":0.7,"pith_summary":"The paper claims that a Borel equivalence relation or groupoid whose equivalence classes carry uniformly Borel (quasi-)Polish topologies is never more expressive than a Polish group action. Its main theorem shows that every Borel-overt fiberwise quasi-Polish groupoid—a standard Borel groupoid with a uniformly Borel family of quasi-Polish topologies on each source fiber—admits a Borel equivalence of groupoids to an open Polish groupoid, and hence to the action groupoid of a Polish group action. This matters because it turns the Becker–Kechris theorem around: instead of starting from a Polish group action, any abstract collection of 'componentwise' Polish topologies satisfying mild uniform-Borelness axioms is realized by a genuine global action. The paper also develops fiberwise versions of Vaught transforms, Effros's theorem on orbits, and the open mapping theorem for these componentwise groupoids.","feed_headline":"Every Borel-overt fiberwise Polish groupoid is a group action","feed_subtitle":"The main theorem turns every abstract locally Polish groupoid into the orbit structure of a Polish group action, up to Borel bireduction.","key_machinery":"The carrying object is the Borel-overt fiberwise quasi-Polish groupoid: a standard Borel groupoid with a uniformly Borel family of quasi-Polish topologies on each fiber of the source map, invariant under right translation, with Borel-testable nonemptiness of fiberwise open sets (the 'overt' condition). The argument proceeds through the componentwise $\\sigma$-topologies $\\mathrm{BO}_{\\mathcal{G}}(X)$ and $\\mathrm{BO}_{\\mathcal{G}}(G)$ on objects and morphisms, which Theorem 4.1.3 shows form an open $\\sigma$-topological groupoid with compatible $\\sigma$-topologies. Two further mechanisms carry the representation: a groupoid-level version of the argument from [SS97] (Lemma 4.3.1) produces a componentwise comeager full subgroupoid on which the difference map $(g,h)\\mapsto g^{-1}h$ is uniformly fiberwise continuous, and a fundamental-sequence criterion from [Ram90] upgrades the resulting quasi-Polish groupoid to a Polish one.","core_discovery":"The central discovery is that the topological information encoded in the Borel structure of a Polish group action can be axiomatized without the action: a Borel-overt fiberwise quasi-Polish groupoid carries exactly the data that a global open Polish groupoid, and ultimately a Polish group action, can produce. On a comeager set of objects the paper constructs a full subgroupoid on which the fiberwise difference operation $(g,h)\\mapsto g^{-1}h$ is uniformly continuous in a Borel way; this uniform condition lets the componentwise $\\sigma$-topologies be assembled into compatible global quasi-Polish topologies. A further comeager pass and a fundamental-sequence criterion upgrade the topology to Polish while preserving the equivalence of groupoids. Consequently every such groupoid is Borel equivalent to the action groupoid of a Polish group action, and the induced equivalence relations are Borel bireducible.","pith_inferences":["Beyond the paper: if the open question whether uniform and non-uniform fiberwise continuity of differences coincide is resolved positively, then classwise quasi-Polish equivalence relations automatically satisfy the uniform hypothesis, and Corollary 4.4.12 could likely be upgraded to a classwise topological embedding into a free Polish group action, as the paper notes in Section 1.6.","Beyond the paper: the idealisticity arguments suggest a definition of 'Borel-idealistic groupoid' (Remark 3.7.7) whose Borel quantifiers preserve Borel sets; the framework makes it plausible that every such groupoid is Borel equivalent to a Polish group action, a groupoid analogue of the conjectured dichotomy for idealistic equivalence relations.","Beyond the paper: the Polishability criterion of Theorem 4.4.5 can be used as a test for whether a concrete fiberwise quasi-Polish groupoid carries a global Polish topology, by checking for symmetric identity-neighborhood sequences inside the componentwise sigma-topology.","Beyond the paper: extending the definitions to analytic equivalence relations with Borel classes, as suggested at the end of the introduction, would give an alternative construction of orbitwise topologies for arbitrary Polish group actions without passing through a groupoid."],"forward_implications":["Every Borel-overt classwise quasi-Polish equivalence relation is Borel bireducible with the orbit equivalence relation of a free Polish group action (Corollary 4.4.12).","Every Borel-overt fiberwise Polish groupoid admits a Borel equivalence of groupoids to an action groupoid of a Polish group action, with Borel inverses-up-to-isomorphism (Corollary 4.4.11 and Proposition 3.8.3).","The componentwise sigma-topologies realize any Borel-overt uniformly componentwise quasi-Polish groupoid as an open quasi-Polish groupoid, with any countably many componentwise open sets included in the compatible topology (Theorem 4.1.5).","The standard toolbox for Polish group actions extends to this abstract setting: Vaught transforms, orbitwise Baire category, idealisticity, Effros's theorem on orbits, the open mapping theorem, and the closed subgroupoid theorem (Sections 3.5-3.7 and 4.5).","When the connectedness relation is smooth, the componentwise topologies assemble into Borel-overt bundles over the quotient, with uniformly open structure maps and Borel sections (Theorem 4.5.5)."],"supporting_citations":[{"why":"Supplies the Becker–Kechris theorem and the motivating example: orbitwise topologies of Polish group actions are exactly the Borel-overt classwise Polish structures.","marker":"[BK96]"},{"why":"Provides the precursor proof and the Borel-bundle, Vaught-transform, and point-free machinery reused throughout.","marker":"[Che24]"},{"why":"Supplies the final step, cited in Corollary 4.4.11, that every open Polish groupoid is Borel equivalent to an action groupoid of a Polish group action.","marker":"[Che19]"},{"why":"Gives the automatic-continuity argument that Lemma 4.3.1 adapts to build subgroupoids with uniformly continuous differences.","marker":"[SS97]"},{"why":"Supplies the fundamental-sequence criterion for identity neighborhoods used to upgrade quasi-Polish groupoids to Polish ones.","marker":"[Ram90]"},{"why":"Defines quasi-Polish spaces and supplies their descriptive-set-theoretic properties used throughout.","marker":"[dB13]"},{"why":"Extends Becker–Kechris and Effros-type results to open Polish groupoids, generalized here to quasi-Polish groupoids.","marker":"[Lup17]"}],"fun_headline_variants":["Componentwise Polish groupoids are all group actions","Fiberwise Polish groupoids become action groupoids","Every Borel-overt Polish groupoid hides a group action","Componentwise Polish topologies force action groupoids"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof needs Lemma 4.3.1's finding of a Borel componentwise comeager full subgroupoid whose fiberwise difference map is uniformly continuous in a Borel way, and the final reduction to a group action also depends on the unpublished result [Che19] that open Polish groupoids are Borel equivalent to Polish group actions.","fun_headline_variants_meta":{"raw":{"variants":["Componentwise Polish groupoids are all group actions","Fiberwise Polish groupoids become action groupoids","Every Borel-overt Polish groupoid hides a group action","Componentwise Polish topologies force action groupoids"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000732,"raw_usage":{"total_tokens":3274,"prompt_tokens":941,"completion_tokens":2333,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":557,"completion_tokens_details":{"reasoning_tokens":2279}},"tokens_in":557,"tokens_out":2333,"duration_ms":19309,"temperature":1.0,"reasoning_tokens":2279,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:55:42.520893+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a Borel-overt fiberwise quasi-Polish groupoid for which no componentwise comeager full subgroupoid has uniformly fiberwise continuous differences, or find a Borel-overt fiberwise Polish groupoid whose connectedness relation is not Borel bireducible with the orbit equivalence relation of any free Polish group action. The paper explicitly leaves open whether uniform and non-uniform continuity of differences are equivalent, so the first is the most direct test.","supporting_citations":[],"review_version":1}