Research

Paper

TESTING March 17, 2026

Neural Pushforward Samplers for the Fokker-Planck Equation on Embedded Riemannian Manifolds

Authors

Andrew Qing He, Wei Cai

Abstract

We extend the Weak Adversarial Neural Pushforward (WANPF) Method to the Fokker--Planck equation posed on a compact, smoothly embedded Riemannian manifold M in $R^n$. The key observation is that the weak formulation of the Fokker--Planck equation, together with the ambient-space representation of the Laplace--Beltrami operator via the tangential projection $P(x)$ and the mean-curvature vector $H(x)$, permits all integrals to be evaluated as expectations over samples lying on M, using test functions defined globally on $R^n$. A neural pushforward map is constrained to map the support of a base distribution into M at all times through a manifold retraction, so that probability conservation and manifold membership are enforced by construction. Adversarial ambient plane-wave test functions are chosen, and their Laplace--Beltrami operators are derived in closed form, enabling autodiff-free, mesh-free training. We present both a steady-state and a time-dependent formulation, derive explicit Laplace--Beltrami formulae for the sphere $S^{n-1}$ and the flat torus $T^n$, and demonstrate the method numerically on a double-well steady-state Fokker--Planck equation on $S^2$.

Metadata

arXiv ID: 2603.16239
Provider: ARXIV
Primary Category: math.NA
Published: 2026-03-17
Fetched: 2026-03-18 06:02

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.16239v1</id>\n    <title>Neural Pushforward Samplers for the Fokker-Planck Equation on Embedded Riemannian Manifolds</title>\n    <updated>2026-03-17T08:24:52Z</updated>\n    <link href='https://arxiv.org/abs/2603.16239v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.16239v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>We extend the Weak Adversarial Neural Pushforward (WANPF) Method to the Fokker--Planck equation posed on a compact, smoothly embedded Riemannian manifold M in $R^n$. The key observation is that the weak formulation of the Fokker--Planck equation, together with the ambient-space representation of the Laplace--Beltrami operator via the tangential projection $P(x)$ and the mean-curvature vector $H(x)$, permits all integrals to be evaluated as expectations over samples lying on M, using test functions defined globally on $R^n$. A neural pushforward map is constrained to map the support of a base distribution into M at all times through a manifold retraction, so that probability conservation and manifold membership are enforced by construction. Adversarial ambient plane-wave test functions are chosen, and their Laplace--Beltrami operators are derived in closed form, enabling autodiff-free, mesh-free training. We present both a steady-state and a time-dependent formulation, derive explicit Laplace--Beltrami formulae for the sphere $S^{n-1}$ and the flat torus $T^n$, and demonstrate the method numerically on a double-well steady-state Fokker--Planck equation on $S^2$.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='math.NA'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <published>2026-03-17T08:24:52Z</published>\n    <arxiv:comment>13 pages, 2 figures, 1 table, 1 algorithm</arxiv:comment>\n    <arxiv:primary_category term='math.NA'/>\n    <author>\n      <name>Andrew Qing He</name>\n    </author>\n    <author>\n      <name>Wei Cai</name>\n    </author>\n  </entry>"
}