Nearby Lagrangian conjecture

About 40 years old · traced to

Let MM be a closed smooth manifold of dimension nn, and let π ⁣:T∗M→M\pi\colon T^{*}M\to M be its cotangent bundle, equipped with the tautological 11-form λ\lambda (in local canonical coordinates (q1,…,qn,p1,…,pn)(q_{1},\dots,q_{n},p_{1},\dots,p_{n}), λ=∑i=1npi dqi\lambda=\sum_{i=1}^{n}p_{i}\,dq_{i}) and the symplectic form ω=dλ\omega=d\lambda.

Call a submanifold L⊂T∗ML\subset T^{*}M Lagrangian if dim⁡L=n\dim L=n and ω∣L=0\omega|_{L}=0, closed if LL is compact without boundary, and exact if there exists f∈C∞(L,R)f\in C^{\infty}(L,\mathbb{R}) with λ∣L=df\lambda|_{L}=df.

Let 0M={(q,0):q∈M}⊂T∗M0_{M}=\{(q,0):q\in M\}\subset T^{*}M denote the zero section. Say that two submanifolds L0,L1⊂T∗ML_{0},L_{1}\subset T^{*}M are Hamiltonian isotopic if there is a smooth family of compactly supported functions Ht ⁣:T∗M→RH_{t}\colon T^{*}M\to\mathbb{R}, t∈[0,1]t\in[0,1], whose time-dependent Hamiltonian vector fields XHtX_{H_{t}}, determined by

ω(XHt,⋅)=dHt,\omega(X_{H_{t}},\cdot)=dH_{t},

generate a flow φt\varphi_{t} with φ0=idT∗M\varphi_{0}=\mathrm{id}_{T^{*}M} and φ1(L0)=L1\varphi_{1}(L_{0})=L_{1}.

Then for every closed exact Lagrangian submanifold L⊂T∗ML\subset T^{*}M, the submanifolds LL and 0M0_{M} are Hamiltonian isotopic.

Equivalent formulations 2Other wordings

Other statements of this same problem, merged from separate entries. Each is equivalent to the statement above — proving any one settles them all.

  1. Nearby Lagrangian conjecture

    In mathematics, more specifically symplectic topology, the nearby Lagrangian conjecture, is an open mathematical problem often attributed to Vladimir Arnold. It states that every closed exact Lagrangian submanifold of a cotangent bundle T∗M is Hamiltonian isotopic to the zero section.

    source: Wikipedia

  2. The nearby Lagrangian conjecture

    Let NN be a closed manifold and let L⊂T∗NL\subset T^*N be an exact Lagrangian, with LL also closed. Nearby Lagrangian conjecture. There exists a Hamiltonian isotopy such that

    L=ϕ(0N).L=\phi(0_N).

    This conjecture is attributed to Arnold and was first published by Lalonde and Sikorav. It has been proved independently by Fukaya–Seidel–Smith and Kragh–Abouzaid.

    source: Marie-Claude Arnaud, Vincent Humilière and Claude Viterbo, “Higher Dimensional Birkhoff attractors (with an appendix by Maxime Zavidovique)”, arXiv:2404.00804 (2026).

References

Primary source

Wikipedia

Additional references

  1. Wikipedia, Nearby Lagrangian conjecture, the article this problem comes from.

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 1

RemarkAI-assistedClaimed by OpenAI. The manuscript claims a closed exact smoothly embedded Lagrangian in the cotangent bundle of S9×SN−1S^9\times S^{N-1}, for some sufficiently large even NN, diffeomorphic to the base but not Hamiltonian isotopic to the zero section.See full solutionHide full solution

Claimed by OpenAI.

Theorem 1.1 claims a counterexample to the unrestricted nearby Lagrangian conjecture: for Q=S9×SN−1Q=S^9\times S^{N-1} with some sufficiently large even NN, a closed exact smoothly embedded Lagrangian L⊂T∗QL\subset T^*Q is diffeomorphic to QQ, but no compactly supported Hamiltonian isotopy carries the zero section to LL. Both the base and the Lagrangian are connected and simply connected. The source and target use opposite signs for the Hamiltonian vector field; replacing the Hamiltonian by its negative leaves the existence of such isotopies unchanged.

GitHub repository: https://github.com/openai/math

Manuscript: https://github.com/openai/math/blob/adc7f1241b42e322a6451854ab7e4b4c146bf78a/preprints/A-counterexample-to-the-nearby-Lagrangian-conjecture-September-23-2026/paper.pdf

  • OpenAI-340-01-A-counterexample-to-the-nearby-Lagrangian-conjecture.pdf303,836 bytesOpen