{"id":"eee45691-a5ce-4b5f-a8ab-f89e48e001d0","arxiv_id":"2502.02935","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A line-bundle approach to contact integrability unifies cooriented and non-cooriented systems and covers dissipative contact Hamiltonians.","lead":"This mathematics paper develops a new geometric framework for integrable dynamical systems on contact manifolds, using line bundles to treat previously separate cases in a unified way. It also extends the theory to dissipative systems, where energy is not preserved, and proves that solutions wind around invariant tori or cylinders.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the flagged rank-drop in Theorem 4(iii) is terse but derivable from Lemma 3(ii); Theorem 3's noncompact linearization follows from the locally free R^{r+1} action.","rationale":"The reader's conditional verdict was motivated by the unproven rank-drop assertion in Theorem 4(iii). On inspection, that assertion is true and can be proved using only statements already in the paper: Theorem 3(i) identifies the fiber tangent space with F_x, the common vanishing of s_0,...,s_r gives F_x subset H_x, and linear algebra gives alpha_i in span{df_j}; Lemma 3(ii) then forces dim E_x = p. Thus the identified weakness is an exposition gap rather than a substantive error. The other potential concern, the noncompact linearization in Theorem 3, is also not an error: complete pointwise-independent commuting vector fields generate a locally free R^{r+1} action, whose orbits are exactly T^l x R^{r+1-l}, with linearized dynamics for any vector field in the span. The paper's use of the author's earlier action-angle theorem [14] is legitimate because [14] is peer-reviewed and provides the normal forms cited. No internal inconsistency or unsupported central step was found. The paper would benefit from a short proof of the rank drop, but this does not affect the correctness of the central claim. Hence the reader's CONDITIONAL verdict can stand, with the condition satisfied by a straightforward elaboration.","tokens_in":20976,"tokens_out":33610,"duration_ms":313001,"concrete_test":"Recompute the rank of E on Sigma in the proof of Theorem 4(iii): on a chart M_{s_i} with i>r, put f_j = s_j/s_i, verify that alpha_i is in span{df_j : j != i} on Sigma, then apply Lemma 3(ii) to the p independent functions {f_j : j != i}; the dimension of span{Z, X_{f_j}} must be p, confirming (4.2).","verdict_should_be":"UNCHANGED","load_bearing_attack":"No load-bearing mathematical error found. The reader's weakest assumption concerns the proof of Theorem 4(iii), where the paper asserts without derivation that on Sigma the rank of E drops from p+1 to p. This assertion is valid and can be derived from existing material in the paper. At x in Sigma, choose i>r with s_i(x) nonzero and set f_j = s_j/s_i on M_{s_i}. By Theorem 3(i), the tangent space of the pi-fiber equals F_x, and since s_0,...,s_r vanish at x, F_x is contained in H_x. Thus alpha_i vanishes on the fiber tangent space, so alpha_i lies in span{df_j : j != i}. Lemma 3(ii) applied to the p independent functions {f_j : j != i} then shows dim span{Z, X_{f_j}} is either p+1 or p; the displayed relation makes p+1 impossible, so dim E_x = p. This proves (4.2) and the claim that Sigma is a singular leaf of E. The proof in the paper is compressed, but the assertion is not circular and does not rely on an unstated assumption. Similarly, Theorem 3's conclusion that connected components are T^l x R^{r+1-l} with linear dynamics follows from completeness and pointwise independence of X_0,...,X_r, which generate a locally free R^{r+1} action with discrete isotropy; no compactness is required. The reliance on the published action-angle results of [14] is a reliance on prior peer-reviewed work, not an internal gap. I therefore find no load-bearing concern that would invalidate or seriously weaken the central claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a line-bundle formulation of contact Hamiltonian systems on a contact manifold (M,H), where the contact structure defines a line bundle L and contact vector fields are identified with sections of L carrying a Jacobi bracket. Given a finite-dimensional space Y of symmetries of a contact system, the paper constructs a momentum map pi: M\\setminus M0 -> RP^p from the sections and proves that the Hamiltonian vector fields of the commuting subalgebra X generate the fibers of pi on regular sets. Theorem 3 states that regular components of invariant level sets are diffeomorphic to T^l x R^{r+1-l} with linearized dynamics, and regular components of the zero locus are T^l x R^{r-l}. Theorem 4 describes the induced foliations F subset E, identifies the singular set Sigma where F is tangent to the contact distribution, and proves that Sigma is a singular leaf of E. Section 5 specializes to commutative integrability on cooriented manifolds, and two examples on the projectivized cotangent bundle of the torus illustrate compact and noncompact invariant manifolds.","tokens_in":21277,"tokens_out":15541,"duration_ms":177148,"significance":"If the results are correct, the paper gives a useful unified geometric framework for contact non-commutative integrability that covers cooriented and non-cooriented contact manifolds, including dissipative systems. The identification of contact symmetries with sections of a contact line bundle and the projective momentum map are natural and potentially valuable constructions. The comparison of the author's pre-isotropic framework [14] with Zung's toric contact integrability is a genuine contribution, and the examples in Section 4 and 5 are concrete and illuminating. The paper is not fully self-contained because it imports the normal-form theorems of [14] and [32], but this reliance is on published results and is not circular. No machine-checked proofs are supplied, but the arguments are standard geometric ones and the main claims appear sound after filling in some omitted details.","major_comments":[{"comment":"The proof asserts without derivation that at points of Sigma the dimension of E drops from p+1 to p, and that this equality characterizes Sigma as in equation (4.2). This rank drop is load-bearing for the singular-leaf conclusion, and the proof as written only shows that the vector fields X_{s0},...,X_{sp} are tangent to Sigma, which alone gives only dim E_x ≤ p. The missing argument can be supplied by applying Lemma 3(ii) on M_{s_i} to the functions f_j = s_j/s_i and using the fact that the momentum map has rank p; this derivation should be written out explicitly in the revised version.","section":"Section 4, Theorem 4(iii)"},{"comment":"The proof of the linearization statement is compressed into a citation to the Arnold-Liouville theorem. In particular, the diffeomorphism type T^l x R^{r+1-l} for possibly noncompact components requires the argument that completeness and pointwise independence of X_0,...,X_r generate a locally free R^{r+1} action, so that each connected component is a quotient of R^{r+1} by a closed subgroup; similarly, on M_{0,reg} the full abelian action has rank r. Since these facts are not stated, the theorem's proof is not fully self-contained and should be expanded.","section":"Section 3, Theorem 3(i)-(ii)"}],"minor_comments":[{"comment":"There is a typographical error in the display after the case dim K_x = r-1: 'a_1 f_1(x) + ... + a_r f(x)' should read 'a_1 f_1(x) + ... + a_r f_r(x)'.","section":"Section 3.6, Lemma 3(ii)"},{"comment":"The phrase 'Sigma is folded on invariant (r+1)-dimensional tori' should read 'Sigma is foliated by invariant (r+1)-dimensional tori'.","section":"Section 4.1"},{"comment":"The notation M0 for the zero locus and M^0_c for connected components of invariant level sets is potentially confusing; using different symbols for these two objects would improve readability.","section":"Section 3.2 and Section 3.4"},{"comment":"The identification L|_{M\\setminus M0} = pi^*O(1) is stated briefly; since it is used in the examples, a short verification of the transition-function computation would be helpful.","section":"Section 3.3"},{"comment":"The wording 'foliated on invariant tori T' should be 'foliated by invariant tori T'.","section":"Corollary 1"}],"recommendation":"major_revision","confidential_remarks":"The paper depends substantially on the author's prior work [14] for action-angle and normal-form results. This is not a circularity problem, but the introduction and the proofs should state explicitly which statements are imported from [14] rather than proved here. The two major comments above concern omitted derivations rather than false claims; once those derivations are added, the paper would be acceptable. The paper fits the scope of the journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Solid, honest paper, and the reader's conditional verdict is fair — maybe slightly conservative, since the flagged gap in Theorem 4(iii) is a missing derivation, not a missing fact. The genuinely new thing is the unified framework: contact integrability formulated on sections of the contact line bundle, with the momentum map π: M ∖ M0 → RP^p, x ↦ [s_0(x):...:s_p(x)], and the identification L restricted to M ∖ M0 with π^*O(1). It covers cooriented and non-cooriented manifolds in one stroke, and it lets the Hamiltonian be dissipative rather than a Reeb integral. Example 3, a dissipative system on T^*T^n × S^1, is concrete and checkable.\n\nThe main theorems hold up. Theorem 3's non-compact linearization (leaves T^l × R^{r+1-l}) follows from completeness: the commuting fields generate a locally free R^{r+1}-action with discrete isotropy, so each leaf is R^{r+1}/Z^l; no compactness is needed. I also checked Σ in Example 2 and the zero locus in Example 3; both match the theorems.\n\nThe soft spots are real but minor. The rank-drop claim in Theorem 4(iii) — dim E_x = p on Σ — is asserted in one sentence. I agree with the stress-test that it is derivable from Lemma 3(ii): at x ∈ Σ pick i > r with s_i(x) ≠ 0; the π-fiber at x lies in H (all s_0,...,s_r vanish there), so α_i annihilates the fiber and lies in span{df_j : j ≠ i}; restricting to H gives a nontrivial relation among the horizontal parts, forcing dim E_x = p. The paper does not give this argument, and a referee should ask for it. Same for Theorem 3(i), which appeals to Arnold–Liouville and [14] where a short paragraph on the locally free action would be cleaner.\n\nThe heavy use of the author's own [14] for action-angle coordinates is a real dependency, but it is published work and not equivalent to the new claims; I do not see circularity. The citation pattern is otherwise sound.\n\nWho this is for: people in contact integrable systems and geometric mechanics. It deserves a serious referee. Send it to review; the revision should expand the two terse proofs, but the framework, theorems, and examples are worth publishing.","headline":"Solid unification of Zung's and Jovanović's contact integrability frameworks via the contact line bundle; the main theorems check out, but two terse proofs need expansion.","tokens_in":21824,"tokens_out":21672,"would_cite":true,"duration_ms":161947,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37J35","37J55","53D10","53C12"],"pacs":[],"model":"deepseek-v4-flash","headline":"A Jacobi bracket on contact line-bundle sections unifies contact integrability across cooriented and non-cooriented manifolds.","keywords":["contact line bundle","Jacobi structure","non-commutative integrability","contact Hamiltonian systems","action-angle variables","pre-isotropic foliation","toric contact integrability","dissipative contact systems"],"falsifier":"Take any system satisfying the hypotheses of Theorem 3 (for instance the n=1 case of Example 2) and compute, in local canonical coordinates, the rank at a regular point of Sigma of the distribution E=span{X_{s0},...,X_{sp}}. Theorem 4(iii) predicts dim E=p there and p+1 just outside Sigma; one point where dim E=p+1 on Sigma would falsify the singular-leaf structure.","tokens_in":20749,"feed_emoji":"🌀","tokens_out":10142,"duration_ms":95984,"temperature":0.7,"pith_summary":"The paper establishes a unified integrability criterion for contact Hamiltonian systems: if a system has sufficiently many pairwise commuting contact symmetries, viewed as sections of the contact line bundle, then its regular invariant level sets are diffeomorphic to a torus times Euclidean space and the flow linearizes on them. The same statement holds on the zero locus of the symmetries, with one fewer dimension, so invariant submanifolds are classified as well. Because the symmetries are sections of the line bundle rather than functions on the manifold, the result covers cooriented and non-cooriented contact manifolds simultaneously, including dissipative Hamiltonians that are not preserved by the original Reeb vector field. The paper also shows how the pre-isotropic noncommutative integrability framework and the toric contact integrability framework fit as pieces of one foliation picture.","feed_headline":"Commuting contact symmetries turn level sets into tori times lines","feed_subtitle":"A Jacobi-bracket framework unifies cooriented and non-cooriented contact manifolds, covering dissipative Hamiltonians.","key_machinery":"The central object is the contact line bundle $L$, built from the local contact forms $\\alpha_U$ with transition functions $g_{UV}$; its smooth sections are identified with contact vector fields through $s=\\alpha(X_s)$, and the Jacobi bracket $[s_1,s_2]=\\Phi([X_{s_1},X_{s_2}])$ makes $\\Gamma(L)$ a Lie algebra. The proof runs through two constructions. On the open set where a section $s$ is nonzero, the ratio $\\phi_s(l)=l/s$ converts arbitrary sections into ordinary functions and makes $X_s$ the Reeb vector field of the contact form $\\alpha_s=\\alpha/s$. The momentum map $\\pi=[s_0:\\dots:s_p]$ into $\\mathbb{RP}^p$ then has invariant fibers, and on each fiber the commuting symmetries act as translations, so the standard torus-and-line linearization applies.","core_discovery":"On a contact manifold $(M,\\mathcal H)$ with associated line bundle $L$, a section $s\\in\\Gamma(L)$ corresponds to a contact vector field $X_s$ by $s=\\alpha(X_s)$, and the Jacobi bracket on sections measures commutativity. The paper's central claim is Theorem 3: given symmetries $s_0=h,\\dots,s_p$ with $[s_i,s_j]=0$ for $i=0,\\dots,r$, $j=0,\\dots,p$, $p+r=2n$, and complete vector fields $X_0,\\dots,X_r$, the momentum map $\\pi:M\\setminus M_0\\to\\mathbb{RP}^p$, $\\pi(x)=[s_0(x):\\dots:s_p(x)]$, has invariant fibers whose regular connected components are diffeomorphic to $T^l\\times\\mathbb{R}^{r+1-l}$ with linearized flow; regular components of the zero locus $M_0$ are $T^l\\times\\mathbb{R}^{r-l}$ with linearized flow. The supporting structure is a flag of distributions $F\\subset E$, where $F$ is spanned by the commuting symmetries and $E$ by all symmetries; dimensions are $p+1$ and $r+1$ on the regular set off a subvariety $\\Sigma$, with $E$ dropping to dimension $p$ on $\\Sigma$ and $F\\subset E\\subset H$ on $M_0$.","pith_inferences":["The paper leaves open the promised treatment of contact reductions and contact dual pairs; if Theorem 3 extends there, integrability of reduced contact systems would follow from symmetry sections of the unreduced system.","The singular leaf $\\Sigma$ is the place where the fibers touch the contact distribution; tracking how invariant tori enter $\\Sigma$ suggests contact analogues of monodromy or scattering, which the paper does not explore.","Because the ratio construction only needs a section to be nonzero locally, the same line-bundle formalism should transplant to other first-order geometries, such as Jacobi or cosymplectic manifolds, where a distinguished vector field is unavailable.","The dependence of invariant topological type on the function $f$ in Example 3 (tori versus non-compact pre-Legendrian manifolds) gives a concrete family in which the singular structure can be tested numerically."],"forward_implications":["Integrability of a contact system is decidable from a set of commuting symmetries alone: with $p+r=2n$, the level sets of the momentum map split into tori and Euclidean factors, and the flow is linear in explicit coordinates.","The zero locus of the symmetries carries the same conclusion in dimension one less, so invariant submanifolds of codimension $p$ are also classified and linearized, not just generic level sets.","Contact action-angle coordinates exist in toroidal neighborhoods of regular invariant tori, with contact normal forms of the types given in equations (2.11) and (4.4).","The statement applies to non-cooriented contact manifolds and to dissipative contact Hamiltonians; the familiar cooriented case is the special case of a trivial line bundle."],"supporting_citations":[{"why":"Supplies the standard linearization of commuting complete vector fields that Theorem 3 applies on each level set.","marker":"[1]"},{"why":"Defines complete pre-isotropic contact structures and the noncommutative integrability theorem that Theorem 4(i) invokes.","marker":"[14]"},{"why":"Provides toric contact integrability and the normal form (4.4) used for the invariant tori in Sigma.","marker":"[32]"},{"why":"Gives the line-bundle construction of the momentum map pi=[s0:...:sp].","marker":"[11]"},{"why":"Supplies the Jacobi bracket and contact-Hamiltonian formalism used to define symmetries as sections.","marker":"[24]"},{"why":"The original noncommutative action-angle formulation that the paper generalizes to contact line bundles.","marker":"[27]"},{"why":"Gives the commutative contact integrability case that the paper recovers when r=n and Sigma is empty.","marker":"[2]"},{"why":"Motivates the dissipative Hamiltonians covered by the cooriented formulation.","marker":"[12]"}],"fun_headline_variants":["Contact symmetries commute: level sets become tori times lines","Jacobi brackets on contact line bundles unify integrable systems","Dissipative contact Hamiltonians fit Jacobi-bracket integrability","Contact line bundles: commuting symmetries produce tori times lines"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"At points where the first set of symmetry sections all vanish, the paper needs the span of all the symmetry vector fields to drop in dimension by exactly one; this rank drop is stated without a derivation, and the claimed singular-leaf description rests on it.","fun_headline_variants_meta":{"raw":{"variants":["Contact symmetries commute: level sets become tori times lines","Jacobi brackets on contact line bundles unify integrable systems","Dissipative contact Hamiltonians fit Jacobi-bracket integrability","Contact line bundles: commuting symmetries produce tori times lines"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000694,"raw_usage":{"total_tokens":3147,"prompt_tokens":962,"completion_tokens":2185,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":578,"completion_tokens_details":{"reasoning_tokens":2113}},"tokens_in":578,"tokens_out":2185,"duration_ms":15235,"temperature":1.0,"reasoning_tokens":2113,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T10:34:09.506940+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take any system satisfying the hypotheses of Theorem 3 (for instance the n=1 case of Example 2) and compute, in local canonical coordinates, the rank at a regular point of Sigma of the distribution E=span{X_{s0},...,X_{sp}}. Theorem 4(iii) predicts dim E=p there and p+1 just outside Sigma; one point where dim E=p+1 on Sigma would falsify the singular-leaf structure.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the standard linearization of commuting complete vector fields that Theorem 3 applies on each level set."},{"cited_title":"Noncommutative integrability and action-angle variables in contact geometry","cited_arxiv_id":"1103.3611","evidence_quote":"Defines complete pre-isotropic contact structures and the noncommutative integrability theorem that Theorem 4(i) invokes."},{"cited_title":"A conceptual approach to the problem of action-angle variables","cited_arxiv_id":"1706.08859","evidence_quote":"Provides toric contact integrability and the normal form (4.4) used for the invariant tori in Sigma."},{"cited_title":"Griffiths, J","cited_arxiv_id":null,"evidence_quote":"Gives the line-bundle construction of the momentum map pi=[s0:...:sp]."},{"cited_title":"Libermann, C","cited_arxiv_id":null,"evidence_quote":"Supplies the Jacobi bracket and contact-Hamiltonian formalism used to define symmetries as sections."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The original noncommutative action-angle formulation that the paper generalizes to contact line bundles."},{"cited_title":"Banyaga, P","cited_arxiv_id":null,"evidence_quote":"Gives the commutative contact integrability case that the paper recovers when r=n and Sigma is empty."},{"cited_title":"Infinitesimal symmetries in Contact Hamiltonian systems","cited_arxiv_id":"1909.07892","evidence_quote":"Motivates the dissipative Hamiltonians covered by the cooriented formulation."}],"review_version":1}