{"id":"584af7aa-4eb8-4dfd-8701-37f2df617bcd","arxiv_id":"2411.16631","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"high","formal_verification":"none","parameter_count":1,"one_line_summary":"On co-adjoint Lie groupoids of product form, Hamiltonian integrability is claimed to reduce to the integrability of a Hamiltonian system on a co-adjoint orbit, but the derivation is not sound.","lead":"This paper claims that Hamiltonian systems on co-adjoint Lie groupoids are integrable exactly when the corresponding system on a co-adjoint orbit is integrable, and that co-adjoint groupoids of symplectic groupoids are again symplectic. Reading it is a way to test whether these structural properties extend from Lie groups to Lie groupoids, though the proof here is largely a citation to the authors' own earlier work.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 23's proof is internally inconsistent: its equations (9) contradict the paper's own formula (7), so the x-independence of H and the reduction in Corollary 27 are unsupported.","rationale":"The paper aims to reduce integrability on a product-form co-adjoint Lie groupoid to integrability on a group co-adjoint orbit. For this reduction to hold, one needs a correct description of the Hamiltonian dynamics on A*\\bar G = M × T*_{\\xi'}O(\\xi') and an iff correspondence of first integrals. The weakest point is not the product restriction per se, since the paper explicitly restricts to O(ξ) = M × O(ξ'), but the proof of Lemma 23, which is the sole bridge to Proposition 26 and Corollary 27. The displayed local formulas are inconsistent with the paper's own general formula (7): the x-equation is wrong, the term '1/2 ∂/∂x^i ∧ ∂/∂x^i' is identically zero, and the conclusion that H is x-independent does not follow from the formulas shown. Moreover, even a repaired product computation likely leaves x as a Casimir direction, in which case x-dependent Hamiltonians have the same orbit dynamics and x itself is an extra first integral, so the 'iff' in Corollary 27 can fail. Since the reduction claim is the central novelty and it rests on this invalid step, the reader's rejection is justified. I do not see a need to change the verdict; the paper would need a corrected Lemma 23 with a full derivation from (5)–(7), and a statement of integrability that accounts for Casimir directions, before the reduction can be assessed.","tokens_in":13971,"tokens_out":7309,"duration_ms":70759,"concrete_test":"Work out the trivial-groupoid example in Section 4 with M = R, G = SU(2), so O(ξ') is a 2-sphere. Compute the linear Poisson tensor on A*\\bar G = R × T*_{\\xi'}S^2 by substituting the structure functions from that example into formula (5), then evaluate X_H for H(x, y) = x + h(y). Check whether (i) the displayed Hamiltonian equations (9) match the correct equations (7); (ii) ∂H/∂x = 1 is compatible with X_H having the same y-component as X_h. If x is a Casimir, the x-dependence is invisible in the orbit dynamics, contradicting Lemma 23. This directly tests the step on which Corollary 27 rests.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central reduction (Corollary 27) rests entirely on Lemma 23, whose proof does not work. In the product case A*\\bar G = M × T*_{\\xi'}O(\\xi'), the general Hamiltonian equations (7) give dx^i/dt = (∂H/∂y_α)ρ^i_α and dy_α/dt = −(∂H/∂x^i)ρ^i_α − (∂H/∂y_β)C^γ_{αβ} y_γ. Lemma 23 instead writes dx^i/dt = ∂H/∂x^i and omits the ∂H/∂x^i term from the y-equation, after displaying a Poisson tensor containing '1/2 ∂/∂x^i ∧ ∂/∂x^i', which is identically zero. This is not a sign slip; the asserted equations are not a specialization of (7) for any structure functions. Consequently, the comparison of (9) with (10) establishes only that, for functions pulled back from T*O(\\xi'), the two Hamiltonian vector fields have the same y-component. It does not force ∂H/∂x^i = 0. In fact, in a product Poisson structure with no x–y cross terms, functions of x may be Casimirs, so H can depend on x while the dynamics on the orbit is governed by h(λ) = H(x_0, λ); such an x-dependent Hamiltonian would add x as a first integral and can make Corollary 27's 'if' direction fail. Thus the integrability reduction is not established by the text.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript continues the authors' earlier work on co-adjoint Lie groupoids. It compares the structure functions of a Lie algebroid and its co-adjoint algebroid (Lemma 22), then, for co-adjoint groupoids of the special product form O(\\xi)=M\\times O(\\xi'), attempts to prove that the Hamiltonian is independent of the base M (Lemma 23), that first integrals correspond to those on O(\\xi') (Proposition 26), and that integrability reduces accordingly (Corollary 27). It also claims that the co-adjoint Lie groupoid of a symplectic groupoid is a symplectic groupoid (Proposition 30).","tokens_in":14399,"tokens_out":16030,"duration_ms":142211,"significance":"If the reduction were correct, it would connect Hamiltonian dynamics on co-adjoint Lie groupoids to the classical Kirillov orbit picture, which would be a useful result. The comparison of structural functions in Section 4 and the three worked examples are potentially helpful. However, the proof of the central Lemma 23 is internally inconsistent with the paper's own general formula, Proposition 26 only treats a restricted class of functions, and Proposition 30 is not proved. The paper also relies heavily on [4] for the groupoid structure on O(\\xi), the bracket identity, and the Poisson bracket equality used in the reduction. As it stands, the announced results are not established.","major_comments":[{"comment":"The Hamiltonian equations displayed in Lemma 23 do not follow from the paper's own general formula (7). In (7) the x-component is (\\partial H/\\partial y_\\alpha)\\rho^i_\\alpha, whereas (9) states dx^i/dt = \\partial H/\\partial x^i. The proof also writes the Poisson tensor as (1/2)\\partial/\\partial x^i \\wedge \\partial/\\partial x^i - (1/2)C^\\gamma_{\\alpha\\beta} y_\\gamma \\partial/\\partial y_\\alpha \\wedge \\partial/\\partial y_\\beta; the first term is identically zero and is incompatible with the first term \\rho^i_\\alpha \\partial/\\partial x^i \\wedge \\partial/\\partial y_\\alpha of the general bivector (5). The y-equation in (9) omits the term -(\\partial H/\\partial x^i)\\rho^i_\\alpha that appears in (7). Hence the conclusion \\partial H/\\partial x^i=0 is not proved.","section":"Section 5.1, Lemma 23 (Eqs. (7), (9))"},{"comment":"Proposition 26 concerns only functions of the special form F=(p,f), meaning functions constant along the M factor. Remark 25 asserts without proof that every smooth function on M\\times T^*_{\\xi'}O(\\xi') has this form, which is false in general. Consequently the converse direction of Proposition 26 applies only to M-independent functions; the paper does not rule out first integrals that genuinely depend on the base coordinates x^i, and for such functions the equality {F,H}_{A^*G}={f,h}_{K.K} used in the proof is not justified.","section":"Section 5.1, Remark 25 and Proposition 26"},{"comment":"Corollary 27 is not established because it assumes H(p,\\lambda)=h(\\lambda), which was supposed to follow from Lemma 23. If H depends on x^i, then the fiber equation from (7) contains the additional coupling term -(\\partial H/\\partial x^i)\\rho^i_\\alpha, so integrability of the orbit Hamiltonian h would not control the full system. Thus the reduction to O(\\xi') is conditional on an unproved and, as written, incorrectly derived independence statement.","section":"Section 5, Corollary 27"},{"comment":"The 2-form \\omega'(\\eta_1,\\eta_2)=\\omega(X,Y) is not shown to be well-defined: the map X\\mapsto ad^*_X\\xi has kernel in general, and no argument is given that \\omega(X,Y) is independent of the choice of representatives X,Y. The proof of multiplicativity is dimensionally incorrect, since it writes m'^*\\omega'(\\eta_1,\\eta_2)=\\omega'(Tm'(\\eta_1,\\eta_2)), evaluating a 2-form on a single tangent vector; a pullback 2-form must be evaluated on a pair of tangent vectors. Closedness and nondegeneracy of \\omega' are never verified, so the proposition does not prove that O(\\xi) is a symplectic groupoid.","section":"Section 6, Proposition 30"},{"comment":"Several load-bearing statements are quoted from [4] without proof: the groupoid structure on O(\\xi) (Theorem 13), the bracket identity [|X',Y'|]'=ad^*_{[|X,Y|]}\\xi (Lemma 18), and the equality {F,H}_{A^*G}={f,h}_{K.K} used in Lemma 23. The last equality, in particular, is used to prove the x-independence of H, but the hypotheses under which it holds are not stated in this manuscript. A continuation note may legitimately cite previous work, but here the quoted result carries the full weight of the main reduction.","section":"Sections 3 and 5, reliance on [4]"}],"minor_comments":[{"comment":"The word 'integrabiliy' in the title should be 'integrability'.","section":"Title"},{"comment":"The abstract contains a spacing typo, 'co-adjo int Lie groupoids'.","section":"Abstract"},{"comment":"The phrase 'if only if' should be 'if and only if'.","section":"Corollary 27"},{"comment":"The statement that H is 'equal to Hamiltonian h' is imprecise; the intended claim is H(p,\\lambda)=h(\\lambda) for all p\\in M and \\lambda\\in T^*_{\\xi'}O(\\xi').","section":"Lemma 23 statement"},{"comment":"In the first example, the basis {e^\\alpha_i} and the formula [|e^\\alpha_i,e^\\beta_j|]=\\theta^\\gamma_{\\alpha\\beta}\\vartheta_\\gamma are not written as an expansion in the same basis; the index ranges and the role of the tangent-bundle summand should be clarified.","section":"Section 4, Examples"},{"comment":"The notion of 'integrable' is not defined in this paper; the reader must infer the definition from references [1,6,14].","section":"Section 5, Corollary 27"},{"comment":"The standing assumption that G is regular is not explicitly used in the statements of Lemma 23 or Proposition 30, so its role should be clarified.","section":"Remark 9"}],"recommendation":"reject","confidential_remarks":"The paper is a short continuation of the authors' work [4] and would need substantial rewriting before it could be considered. The flaw in Lemma 23 is not a local typo: the displayed equations are inconsistent with equation (7). In addition, the main reduction is narrower than the abstract suggests, and Proposition 30 lacks a coherent proof. I recommend rejection rather than major revision because the central claim is not supported by the text as written."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here is my take. The two headline claims are not established. Lemma 23, which is the load-bearing step for the integrability reduction, has a proof that contradicts the paper's own general formula (7). The Hamiltonian equations (9) in Lemma 23 are not a specialization of (7) for any structure functions: the dx/dt equation should involve ∂H/∂y_α, not ∂H/∂x^i, and the dy/dt equation should contain the missing ∂H/∂x^i term. The Poisson bivector displayed in the proof includes '1/2 ∂/∂x^i ∧ ∂/∂x^i', which is identically zero. So the proof is internally inconsistent, and the conclusion that H is independent of the base coordinates is unsupported. Corollary 27 rests entirely on that lemma, so the claimed reduction is at best conditional on an assumption the paper fails to prove.\n\nWhat is genuinely new: Proposition 30, stating that the co-adjoint groupoid of a symplectic groupoid is symplectic, is not in the cited references, as far as I can tell. The paper also frames the integrability reduction cleanly, and the worked examples for trivial, gauge, and action groupoids are standard but useful. The authors are honest that this is a continuation of [4], and they cite their earlier work clearly.\n\nWhere the paper falls short beyond Lemma 23: Proposition 30 defines ω'(η1,η2)=ω(X,Y) without checking well-definedness. The map X↦ad*_X ξ has a kernel in general, so if ad*_X ξ=ad*_{X'}ξ, one must verify ω(X,Y)=ω(X',Y) before calling ω' a form. No such check appears. The multiplicativity argument also mishandles the pullback notation, writing ω'(ad*_X ξ) as if ω' took one argument. So the proof of Prop 30 is missing a real argument. Nearly all foundational statements are delegated to [4], which is acceptable for a continuation but leaves this note with little independent content.\n\nWho is this for? Readers already working with the authors' co-adjoint groupoids who want the hoped-for reduction stated compactly. As written, I would desk-reject it. The authors should fix Lemma 23 by deriving the Hamiltonian equations correctly from (7), restate Corollary 27 under the explicit hypothesis H(p,λ)=h(λ) if that is what they need, and supply a genuine well-definedness proof for the symplectic form in Prop 30. If those repairs are made, the note could become useful, but it does not deserve referee time in its current form.","headline":"The integrability reduction and symplectic inheritance claims are not proven; Lemma 23's proof contradicts the paper's own equations.","tokens_in":14878,"tokens_out":7869,"would_cite":false,"duration_ms":71041,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18B40","53D17","70H08","37J35"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that, for co-adjoint Lie groupoids of product form, integrability of Hamiltonian systems on the groupoid side is equivalent to integrability on the corresponding co-adjoint orbit of a Lie group.","keywords":["co-adjoint Lie groupoid","co-adjoint Lie algebroid","Hamiltonian system","integrable Hamiltonian system","symplectic Lie groupoid","linear Poisson structure","structure functions"],"falsifier":"Construct a regular Lie groupoid whose co-adjoint orbit is not diffeomorphic to a product M × O(ξ′), choose a Hamiltonian on A*G that varies along M, and check whether its first integrals still correspond to those of a Hamiltonian on O(ξ′); a single such example with non-corresponding first integrals would break Corollary 27.","tokens_in":13724,"feed_emoji":"⚙️","tokens_out":5435,"duration_ms":51399,"temperature":0.7,"pith_summary":"The paper studies Hamiltonian systems on co-adjoint Lie groupoids, which are groupoids built from orbits of the co-adjoint action of a Lie groupoid on its isotropy Lie algebroid. Its central claim is that when such a co-adjoint groupoid has the product form O(ξ) = M × O(ξ′), where O(ξ′) is a co-adjoint orbit of a Lie group, integrability of the Hamiltonian system on the groupoid side is exactly the same as integrability of the corresponding system on the orbit O(ξ′). This matters because it reduces an apparently more complicated groupoid-level integrability question to the classical, well-studied integrability of Hamiltonian systems on co-adjoint orbits of Lie groups. The paper also shows that the structure functions of the co-adjoint Lie algebroid coincide with those of the original Lie algebroid, and that the co-adjoint groupoid of a symplectic groupoid is again a symplectic groupoid.","feed_headline":"Integrability on co-adjoint Lie groupoids reduces to Lie-group orbits","feed_subtitle":"For product-type groupoids, first integrals match those on the underlying orbit, so integrability transfers exactly.","key_machinery":"The central device is the co-adjoint Lie groupoid O(ξ) = M × O(ξ′) and its co-adjoint Lie algebroid A*G = M × T*_{ξ′}O(ξ′), together with the linear Poisson structure determined by the original Lie algebroid's structure functions. The argument rides on the identity {F, H}_{A*G} = {f, h}, where the right-hand side is the canonical Lie–Poisson bracket on T*O(ξ′), for Hamiltonians of the form H(p, λ) = h(λ); this identity transfers Hamiltonian vector fields and first integrals from the product groupoid to the Lie-group orbit. Lemma 22's equality of anchor coefficients and bracket constants makes the transfer explicit in local coordinates and is what lets the paper compare the two Hamiltonian systems.","core_discovery":"For a co-adjoint Lie groupoid of the special product form O(ξ) = M × O(ξ′), with O(ξ′) ⊂ g* a co-adjoint orbit of a Lie group, the Hamiltonian dynamics on the dual of the co-adjoint Lie algebroid is exactly the Hamiltonian dynamics on T*O(ξ′) with the M-direction inert. The first integrals of the Hamiltonian vector field on A*G are in one-to-one correspondence with the first integrals of the corresponding vector field on T*O(ξ′), so the Hamiltonian system on the groupoid side is integrable exactly when the system on the Lie-group orbit is integrable. The proof hinges on the equality {F, H}_{A*G} = {f, h}, where F = (p, f), H(p, λ) = h(λ), and the bracket on the right is the canonical Lie–Poisson bracket on the orbit. The paper also establishes that the anchor coefficients and bracket constants of the co-adjoint Lie algebroid equal those of the original Lie algebroid, and that co-adjoint Lie groupoids inherit symplectic groupoid structures from symplectic groupoids.","pith_inferences":["The paper leaves open whether the reduction holds when the co-adjoint orbit is not a product O(ξ) = M × O(ξ′); a natural test is a regular Lie groupoid whose co-adjoint orbit is a nontrivial bundle over M, where the M-direction may no longer be inert.","The structure-function equality suggests that other Poisson-geometric data—such as bi-Hamiltonian chains or Casimir functions—might transfer between a Lie algebroid and its co-adjoint algebroid, though the paper does not pursue that transfer.","The symplectic groupoid result could be relevant to deformation quantization of co-adjoint groupoids, since a multiplicative symplectic form is the standard starting point for such quantizations, but the paper only establishes the symplectic structure itself.","A testable extension would be to weaken the product-form hypothesis by allowing O(ξ) to be a locally trivial fibration over M with typical fiber O(ξ′); the paper gives no evidence for whether Corollary 27 would survive in that generality."],"forward_implications":["If the central claim is correct, integrability of Hamiltonian systems on product-type co-adjoint Lie groupoids is fully equivalent to integrability on the corresponding co-adjoint orbit of a Lie group (Corollary 27).","First integrals on A*G for such groupoids are in bijection with first integrals on T*O(ξ′), so any integrability result for Lie-group orbits immediately carries over to the groupoid setting.","The equality of structure functions (Lemma 22) means the local Poisson data of the co-adjoint Lie algebroid are the same as those of the original Lie algebroid, so computations on either side are interchangeable in the product case.","For the examples treated in the paper—trivial groupoids, gauge groupoids, and transitive action groupoids—the co-adjoint orbits are of the product form, so the reduction applies to each of these families.","If the original Lie groupoid is symplectic, its co-adjoint Lie groupoid is symplectic, which extends the symplectic-groupoid property to this class of co-adjoint constructions."],"supporting_citations":[{"why":"Defines co-adjoint Lie groupoids and co-adjoint Lie algebroids, and supplies the bracket and action relations used in Lemma 22 and Proposition 26.","marker":"[4]"},{"why":"Provides the background definitions of Lie groupoids, Lie algebroids, structure functions, and symplectic groupoids that the paper builds on.","marker":"[10]"},{"why":"Supplies the framework of integrable Hamiltonian systems and first integrals that Proposition 26 and Corollary 27 invoke.","marker":"[1]"},{"why":"Gives an integrable Hamiltonian system on the Lie algebra so(4), serving as an orbit-side model of the systems the reduction targets.","marker":"[6]"},{"why":"Provides a class of quadratic so(4) Hamiltonians, another orbit-side integrable example used for comparison.","marker":"[14]"},{"why":"Establishes that the isotropy groupoid of a regular Lie groupoid is a Lie groupoid, which is needed for co-adjoint orbits to be Lie groupoids.","marker":"[13]"},{"why":"Describes Lie algebroid and Lie groupoid actions and states that co-adjoint orbits of gauge groupoids are of the form P × O(ξ′)/G, used in the examples.","marker":"[2]"}],"fun_headline_variants":["Co-adjoint groupoid integrability collapses to orbit case","Hamiltonian integrability on co-adjoint groupoids equals orbit dynamics","When co-adjoint groupoids split, integrability matches Lie orbits","First integrals on co-adjoint groupoids reduce to group orbits"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The reduction only works when the co-adjoint orbit splits as a product O(ξ) = M × O(ξ′) and the Hamiltonian depends only on the second factor; without that product structure the paper gives no route from co-adjoint groupoid integrability to Lie-group orbit integrability.","fun_headline_variants_meta":{"raw":{"variants":["Co-adjoint groupoid integrability collapses to orbit case","Hamiltonian integrability on co-adjoint groupoids equals orbit dynamics","When co-adjoint groupoids split, integrability matches Lie orbits","First integrals on co-adjoint groupoids reduce to group orbits"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000227,"raw_usage":{"total_tokens":1458,"prompt_tokens":916,"completion_tokens":542,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":532,"completion_tokens_details":{"reasoning_tokens":466}},"tokens_in":532,"tokens_out":542,"duration_ms":5625,"temperature":1.0,"reasoning_tokens":466,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:54:51.302095+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a regular Lie groupoid whose co-adjoint orbit is not diffeomorphic to a product M × O(ξ′), choose a Hamiltonian on A*G that varies along M, and check whether its first integrals still correspond to those of a Hamiltonian on O(ξ′); a single such example with non-corresponding first integrals would break Corollary 27.","supporting_citations":[{"cited_title":"Haghighatdoost and R","cited_arxiv_id":null,"evidence_quote":"Defines co-adjoint Lie groupoids and co-adjoint Lie algebroids, and supplies the bracket and action relations used in Lemma 22 and Proposition 26."},{"cited_title":"Mackenzie, General theory of Lie groupoids and Lie algebroids , London Math","cited_arxiv_id":null,"evidence_quote":"Provides the background definitions of Lie groupoids, Lie algebroids, structure functions, and symplectic groupoids that the paper builds on."},{"cited_title":"Bolsinov and A.T","cited_arxiv_id":null,"evidence_quote":"Supplies the framework of integrable Hamiltonian systems and first integrals that Proposition 26 and Corollary 27 invoke."},{"cited_title":"Haghighatdoost and A.A Oshemkov, The topology of Liouville foliation for the Sokolov integra ble case on the Lie algebra so(4) , Sbornik: Mathematics, 200(6), 899 – 921, (2009)","cited_arxiv_id":null,"evidence_quote":"Gives an integrable Hamiltonian system on the Lie algebra so(4), serving as an orbit-side model of the systems the reduction targets."},{"cited_title":"Sokolov, One class of quadratic so(4) Hamiltonians , Dokl","cited_arxiv_id":null,"evidence_quote":"Provides a class of quadratic so(4) Hamiltonians, another orbit-side integrable example used for comparison."},{"cited_title":"Schmeding, The Lie group of vertical bisections of a regular Lie groupoi d, Forum Mathematicum, 32, 479–489, (2019)","cited_arxiv_id":null,"evidence_quote":"Establishes that the isotropy groupoid of a regular Lie groupoid is a Lie groupoid, which is needed for co-adjoint orbits to be Lie groupoids."},{"cited_title":"Bos, Geometric quantization of Hamiltonian actions of Lie algeb roids and Lie groupoids , Int","cited_arxiv_id":null,"evidence_quote":"Describes Lie algebroid and Lie groupoid actions and states that co-adjoint orbits of gauge groupoids are of the form P × O(ξ′)/G, used in the examples."}],"review_version":1}