{"id":"7dd27fdc-a8b5-42b2-9766-438270b8190e","arxiv_id":"2507.08620","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors develop a deformation theory for coisotropic branes, explicitly describing nearby branes in a codimension-one model and showing that the forgetful map from branes to coisotropic submanifolds is not locally surjective.","lead":"In symplectic geometry, a brane is a coisotropic submanifold equipped with an extra closed 2-form. This paper classifies small deformations of such branes and proves that some coisotropic submanifolds arbitrarily close to a brane carry no brane structure.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The K3-based Theorem 3.17 is credible, but the obstruction map in Remark 4.20 is asserted without proof and used in Example 4.21, so the advertised T^4 obstruction-vanishing claim is conditional.","rationale":"The reader's conditional verdict is reasonable. I agree that the paper should not be accepted outright while a nontrivial new claim is deferred. However, I identify the load-bearing gap slightly differently: the most concrete and self-admitted missing proof is the obstruction map of Remark 4.20, which is used in Example 4.21 and advertised in the introduction. The headline K3 application, by contrast, depends on standard cited facts: the rigidity of automorphisms of complex K3 surfaces and Debord's period bound. Assuming those external results, Proposition 3.14 and Theorem 3.17 are internally coherent. I therefore do not see grounds to reject the paper or to move beyond the reader's CONDITIONAL verdict; the appropriate action is to condition acceptance on supplying the promised DGLA details or on removing the obstruction-vanishing claim from the stated results.","tokens_in":29059,"tokens_out":20409,"duration_ms":256450,"concrete_test":"Complete the deferred DGLA proof for Remark 4.20: verify explicitly that graph(-F) induces the stated DGLA on Ω(M)[1], that the differential is d, the bracket is -1/2 F^{-1}, and that Maurer-Cartan elements inside Ω^{1,1}_R(M) ⊕ ⊕_{k≥3}Ω^k(M) correspond exactly to nearby space-filling brane structures. If this verification succeeds, Example 4.21 is justified; if it cannot be given, delete the obstruction-vanishing claim from Example 4.21 or label it as conditional, and rely only on Lemma 4.22 for the T^4 prolongation statement.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's own limitation statement in Remark 4.20 says 'details and proofs will appear in a separate paper' for the obstruction map defined in (14): B maps to [B,B]_{F^{-1}} modulo dΩ^{1,1}_R(M). This map is then used in Example 4.21 to conclude that on the 4-torus the obstruction map vanishes identically, and the introduction advertises this as a result. That conclusion has no proof in the present text: the DGLA construction via graph(-F), its claimed differential d, its bracket -1/2 F^{-1}, and the identification of Maurer-Cartan elements with brane structures are all deferred. Lemma 4.22 independently proves prolongability of infinitesimal deformations on T^4, but it does not prove that (14) is the correct obstruction or that its vanishing is necessary. The headline application Theorem 3.17 does not depend on Remark 4.20; its reliance on the K3 rigidity theorem [Huy16, BPVdV84] and on Debord's period bound is standard external input, and the logical chain in Proposition 3.14 is sound if those inputs are accepted. The concrete gap is therefore the unproved, newly claimed obstruction map, which makes part of the paper conditional rather than fully established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies coisotropic branes in symplectic manifolds, i.e. coisotropic submanifolds Y equipped with a closed 2-form F whose kernel equals the characteristic distribution and which induces a transverse complex structure. The main results are: (1) for the codimension-one model Y = N × S^1 × {0} in (N × S^1 × R, ω_N × ω_{T^*S^1}), a bijection between brane structures on graph(f) and Φ^1_f-invariant space-filling brane structures on N (Proposition 3.4); (2) an infinitesimal deformation theory for arbitrary branes, encoded in a cochain complex (§4.1–4.2); (3) a proof that there exists a compact brane Y such that arbitrarily C^2-close coisotropic submanifolds need not admit brane structures, using K3 surfaces (Theorem 3.17); and (4) for space-filling branes, an obstruction map for prolonging infinitesimal deformations, which is claimed to vanish on the 4-torus, together with a separate proof that all infinitesimal deformations on T^4 can be prolonged (Lemma 4.22).","tokens_in":29384,"tokens_out":29951,"duration_ms":345102,"significance":"The results, if fully established, are a valuable contribution to the symplectic and generalized-complex geometry of branes. Theorem 3.17 is a strong and interesting negative result: it shows that the forgetful map from branes to coisotropic submanifolds is not locally surjective, and the proof via K3 automorphism rigidity is credible. The explicit infinitesimal deformation complex in §4.2 is a useful new tool, and the T^4 prolongation statement is a concrete positive result. However, the paper advertises an obstruction map for space-filling branes whose construction and properties are deferred to a separate paper; the T^4 vanishing statement for that map is therefore conditional. The main theorems that do not rely on the obstruction map are essentially sound, but the manuscript should not present the obstruction-theoretic claims as established.","major_comments":[{"comment":"The obstruction map (14) is introduced with the sentence \"We claim that if an infinitesimal brane deformation B is not mapped to zero ... then B can not be prolonged,\" and Remark 4.20 explicitly states that \"details and proofs will appear in a separate paper.\" This is a load-bearing point because Example 4.21 uses the map (14) to conclude that on T^4 the obstruction map vanishes identically, and the introduction advertises this as a result. The DGLA construction via graph(-F), the claimed differential and bracket, and the identification of Maurer-Cartan elements with brane structures are not proven here. Lemma 4.22 proves that infinitesimal deformations on T^4 can be prolonged, but it does not establish that (14) is the correct obstruction or that its vanishing is necessary for prolongability. The authors should either provide the proof of the obstruction map and its properties in this paper, or clearly mark the obstruction-theoretic statements in Remark 4.20 and Example 4.21 as conditional/deferred, removing them from the list of established results.","section":"§4.2, Remark 4.20 and Example 4.21"},{"comment":"The proof of Lemma 3.2 is not written out; the paper says it follows immediately from Lemma A.1. Lemma A.1 requires condition (ii) for all vector fields X ∈ Γ(E), whereas Lemma 3.2 only imposes the invariance condition for those X whose flow preserves the complement G. The reduction from all X to this subclass is not explained. In the application in Lemma 3.3 the distributions satisfy the additional property [E_f, G] ⊂ Γ(G), which makes the reduction work, but as stated Lemma 3.2 is not justified by Lemma A.1. The authors should state the missing hypothesis explicitly or supply the missing argument that condition (ii) for the smaller class of vector fields implies (L_X F)|_{∧^2 G} = 0 for all X ∈ Γ(E).","section":"§3.2, Lemma 3.2"}],"minor_comments":[{"comment":"The proof refers to \"Remark 4.19\", but the relevant statement appears to be Lemma 4.19 (or Remark 4.18); please correct the cross-reference.","section":"§A.3, proof of Lemma 4.22"},{"comment":"The notation \"F^{-1}\" is used for the Poisson bivector inverse to the symplectic form F; this is clear from context, but a brief reminder of the convention in the text next to equation (14) would improve readability.","section":"§4.2, Remark 4.20"},{"comment":"The existence of the time-1 flow Φ^1_f is an assumption in Lemma 3.3 and Proposition 3.4; the text notes that compact support guarantees existence, but the statement of Proposition 3.4 does not include the compact-support hypothesis. Please clarify in the statement that the hypothesis is the existence of Φ^1_f.","section":"§3.1, Lemma 3.3"}],"recommendation":"major_revision","confidential_remarks":"The core K3-based theorem and the infinitesimal deformation complex are in good shape, and the deferred obstruction map is the main source of conditional content. I also recommend that the authors re-examine the statement of Lemma 3.2, since the passage from Lemma A.1 to Lemma 3.2 is not immediate as written. The use of the authors' own [KLZ25] for the T^4 prolongation is legitimate and not circular, but the manuscript should be transparent that the obstruction map itself remains a separate-paper result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"That paper is worth engaging with. It makes real progress on the deformation theory of coisotropic branes, and most of the main structural results are proven. The one significant caveat is the obstruction map in Remark 4.20, which is advertised but not proved here; Example 4.21 rests on it. The K3-based Theorem 3.17 does not depend on that obstruction map.\n\nWhat is actually new: the paper goes beyond the infinitesimal, exact-B treatments in [KM06] and [Col14]. Proposition 3.4 gives a clean bijection between brane structures on graph(f) and Phi^1_f-invariant space-filling brane structures on N; the proof via Lemmas 3.2 and 3.3 is self-contained. Corollary 4.7 and the cochain complexes in Propositions 4.16, 4.24, and 4.26 are genuinely new, and the framework handles closed but non-exact B. The negative result on the infinitesimal surjectivity of the forgetful map is concrete and explicit: condition (22) in Proposition 4.32 is usable, and the examples on R^4/T^4 and on K3 make the point convincingly. Theorem 3.17 is credible: if you accept the standard K3 rigidity theorem and Debord's period bounding lemma, the chain from Proposition 3.14 works. Those are legitimate external inputs, not hidden assumptions.\n\nThe soft spot is exactly what the stress-test note identifies. Remark 4.20 says, in the paper's own words, that details and proofs for the obstruction map (14) will appear in a separate paper. That map is then used in Example 4.21 to claim the obstruction vanishes on T^4. Lemma 4.22 proves that infinitesimal deformations on T^4 can be prolonged, but that is a different statement: it does not establish that (14) is the correct obstruction, nor that its vanishing is necessary. So the advertised obstruction result is conditional as written. This is a real gap, but not a fatal one. The central deformation-theoretic content and the K3 application do not rely on it. The repeated self-citations to [KLZ25] are not circular; they import published results, and the main deformation descriptions are derived here from first principles.\n\nWho is this for: people working on generalized complex geometry, coisotropic deformations, or mirror-symmetry-style branes. It deserves a serious referee. My recommendation: send it out, and direct the referee to focus on Section 4.2. Ask the authors to either prove the obstruction statement or visibly mark it as conditional. With that cleaned up, this paper should be published.","headline":"Solid deformation theory for coisotropic branes; the K3-based non-brane result holds, but the advertised T^4 obstruction map is stated without proof and should be conditionalized.","tokens_in":29891,"tokens_out":2472,"would_cite":true,"duration_ms":29932,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53D05","53D17","53D12"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper shows that brane structures on coisotropic submanifolds are genuinely restrictive: even infinitesimally, nearby coisotropic deformations can carry no brane structure, and it supplies the deformation theory and cochain complexes…","keywords":["coisotropic branes","brane structures","symplectic manifolds","deformation theory","coisotropic submanifolds","space-filling branes","K3 surfaces","cochain complexes"],"falsifier":"Search for a non-constant smooth function $g$ on a K3 surface whose Hamiltonian vector field has a time-one flow equal to the identity. If such a function exists (even for large $g$), pulling it back to $Y = N \\times S^1$ would make graph$(f)$ a brane arbitrarily close to $Y$, contradicting Theorem 3.17; the theorem's proof rules this out only by relying on the K3 rigidity statement and the period-bounding lemma, so an explicit example would pinpoint which input fails.","tokens_in":28846,"feed_emoji":"","tokens_out":7882,"duration_ms":75318,"temperature":0.7,"pith_summary":"The paper studies branes—coisotropic submanifolds equipped with a closed 2-form that induces a transverse complex structure—in symplectic manifolds. It establishes two things. First, for a class of codimension-one branes of the form $N \\times S^1$ embedded in $N \\times T^*S^1$, all brane structures on small deformations (graphs of functions) are classified by space-filling brane structures on $N$ that are invariant under a certain diffeomorphism $\\Phi^1_f$. Second, for arbitrary branes, infinitesimal deformations are described as pairs $(r,B)$ satisfying explicit linear equations, governed by a cochain complex. The headline application is that there exists a compact brane $Y$ for which arbitrarily small $C^2$ perturbations are coisotropic yet admit no brane structure, so the forgetful map from branes to coisotropic submanifolds is not surjective near $Y$.","feed_headline":"Close to a brane, coisotropic submanifolds can be brane-free","feed_subtitle":"Small deformations of a brane in a K3 product remain coisotropic but may be brane-free.","key_machinery":"The load-bearing object is the brane structure (Definition 2.4): a closed 2-form $F$ on a coisotropic submanifold $Y$ sharing the constant-rank kernel $E$ with the pullback of the ambient symplectic form, and making $\\omega^{-1}\\circ F$ a complex structure on $TY/E$. Around this, the paper assembles three mechanisms: (i) a bijection (Proposition 3.4, via Lemma 3.3) between brane structures on graph$(f)$ and $\\Phi^1_f$-invariant space-filling brane structures on the symplectic base $N$, built by extending a symplectic form along the characteristic foliation; (ii) a cochain complex (Propositions 4.16 and 4.24) whose degree-1 cocycles are infinitesimal brane deformations and whose coboundaries are Hamiltonian symmetries; and (iii) two rigidity inputs—the K3 surface automorphism rigidity and a period-bounding lemma for Hamiltonian flows—that turn the bijection into a non-existence result for nearby branes.","core_discovery":"The central discovery is that the brane condition is strictly stronger than the coisotropic condition, and that its failure can be measured. For the model $Y = N \\times S^1 \\times \\{0\\}$ in $N \\times T^*S^1$, a graph$(f)$ has a brane structure exactly when $N$ carries a space-filling brane structure preserved by the time-one map $\\Phi^1_f$ of a Hamiltonian flow determined by $f$ (Proposition 3.4). When $N$ is a K3 surface with a space-filling brane, that invariance condition forces $\\Phi^1_f$ to be the identity (using the fact that an automorphism of a complex K3 surface which is trivial on $H^2(X,\\mathbb{Z})$ is trivial), so any non-trivial small Hamiltonian flow produces a coisotropic graph that is arbitrarily close to $Y$ but has no brane structure (Theorem 3.17). Infinitesimally, the paper describes all brane deformations as pairs $(r,B)$ of a closed foliated 1-form and a closed horizontal 2-form satisfying linear equations, presents cochain complexes whose 1-cocycles are these deformations and whose coboundaries are Hamiltonian symmetries, and shows the infinitesimal forgetful map is not surjective in general.","pith_inferences":["Beyond the paper, the explicit obstruction map (14) suggests a deformation-obstruction theory for space-filling branes in higher dimensions: one could test whether vanishing of $[B,B]_{F^{-1}}$ modulo exact $(1,1)$-forms is also sufficient for prolongation, not merely necessary.","The codimension-one bijection hints that nearby branes in topologically non-trivial symplectic manifolds are governed by the dynamics of Hamiltonian diffeomorphisms; a natural testable extension is to replace the circle bundle with higher-rank torus bundles, where the relevant invariance would be under a torus action rather than a single time-one map.","The K3 example gives a compact, concrete counterexample to the intuitive expectation that coisotropic deformations of branes remain branes; a possible use is in generalized complex geometry, where the same graph construction might produce counterexamples to openness of the brane locus in the space of generalized complex submanifolds.","One could compute the full cohomology of the cochain complex (16) for the K3 example and compare it with the moduli space of holomorphic symplectic structures, testing whether the infinitesimal rigidity reflects a finite-dimensional moduli space of nearby branes."],"forward_implications":["For $N$ a K3 surface, every coisotropic submanifold that is $C^2$-close to the product brane and admits a brane structure must equal the graph of a function whose Hamiltonian time-one map is the identity; non-trivial small Hamiltonian perturbations give brane-free coisotropics arbitrarily close to $Y$.","All branes that are $C^1$-close to $Y = N \\times S^1 \\times \\{0\\}$ arise from the mapping torus construction applied to a symplectomorphism of $N$ preserving a space-filling brane structure (Proposition 3.9), so the classification of nearby branes reduces to the symplectic geometry of $N$.","For any brane, the infinitesimal deformations are the closed elements of an explicit subspace of $\\Gamma(E^*) \\oplus \\Omega^2(Y)$, and the cohomology of the associated cochain complex is the formal tangent space to the moduli space of branes modulo Hamiltonian isotopies.","On the 4-torus, every infinitesimal deformation of a space-filling brane prolongs to a 1-parameter family of space-filling branes, and the obstruction map vanishes (Example 4.21, Lemma 4.22).","The infinitesimal forgetful map from brane deformations to coisotropic deformations is not surjective; for the codimension-one model its image is exactly the $r$ satisfying $d_N I^* d_N(\\int_{S^1} r)=0$ (Proposition 4.32), so many coisotropic directions are not brane directions."],"supporting_citations":[{"why":"introduces branes and their characterization through generalized complex geometry, which the paper takes as Definition 2.4.","marker":"[Gua03]"},{"why":"provides the tubular neighborhood theorem that lets every presymplectic manifold sit coisotropically in a symplectic manifold, used throughout.","marker":"[Got82]"},{"why":"supplies the deformation theory of coisotropic submanifolds and the formula $\\omega_f = \\omega_N - d(f\\,dq)$ for graphs.","marker":"[OP05]"},{"why":"states the K3 rigidity used in Proposition 3.16: a complex K3 automorphism trivial on $H^2(X,\\mathbb{Z})$ is the identity.","marker":"[Huy16]"},{"why":"also supplies the K3 automorphism rigidity result invoked in Proposition 3.16.","marker":"[BPVdV84]"},{"why":"provides the period-bounding lemma used in Theorem 3.17 to ensure small nontrivial Hamiltonian flows do not have identity time-one maps.","marker":"[Deb00]"},{"why":"gives the generalized-complex-geometry proof pattern that Lemma 4.3 adapts to derive the linearized brane equations.","marker":"[Col14]"},{"why":"offers earlier infinitesimal brane deformations, which the paper generalizes by allowing non-exact closed 2-forms $B$.","marker":"[KM06]"},{"why":"supplies facts about space-filling branes on symplectic 4-manifolds, including the smoothness of the moduli space and the $T^4$ prolongation argument.","marker":"[KLZ25]"}],"fun_headline_variants":["Arbitrarily close to a brane, coisotropic submanifolds can be brane-free","Near any brane, coisotropic submanifolds may lack brane structures","Coisotropic submanifolds need not be branes, even near branes","Brane-free coisotropics exist arbitrarily close to a brane","Small Hamiltonian flows yield brane-free coisotropic graphs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument stands on an imported rigidity theorem for K3 surfaces: a complex-surface automorphism that acts trivially on integral second cohomology must be the identity. If that theorem admitted exceptions, the paper's compact counterexample would collapse.","fun_headline_variants_meta":{"raw":{"variants":["Arbitrarily close to a brane, coisotropic submanifolds can be brane-free","Near any brane, coisotropic submanifolds may lack brane structures","Coisotropic submanifolds need not be branes, even near branes","Brane-free coisotropics exist arbitrarily close to a brane","Small Hamiltonian flows yield brane-free coisotropic graphs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000798,"raw_usage":{"total_tokens":3488,"prompt_tokens":903,"completion_tokens":2585,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":519,"completion_tokens_details":{"reasoning_tokens":2478}},"tokens_in":519,"tokens_out":2585,"duration_ms":18706,"temperature":1.0,"reasoning_tokens":2478,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:14:17.360003+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for a non-constant smooth function $g$ on a K3 surface whose Hamiltonian vector field has a time-one flow equal to the identity. If such a function exists (even for large $g$), pulling it back to $Y = N \\times S^1$ would make graph$(f)$ a brane arbitrarily close to $Y$, contradicting Theorem 3.17; the theorem's proof rules this out only by relying on the K3 rigidity statement and the period-bounding lemma, so an explicit example would pinpoint which input fails.","supporting_citations":[],"review_version":1}