Energy-Guided Transport for Projection-Free Physics-Informed Flow Matching

Abstract

Deep generative models, particularly Flow Matching frameworks, have emerged as powerful tools for synthesizing complex physical systems. However, standard architectures remain purely statistically driven and agnostic to the underlying governing equations, often violating fundamental physical constraints during inference. Existing zero-shot approaches designed to enforce these constraints typically rely on rigid algebraic projections. While effective for sparse boundary conditions, applying such discrete geometric corrections to dense, full-domain PDEs may disrupt the probability path, \rev{potentially deviating from the learned manifold and introducing substantial computational overhead}. To address these challenges, we introduce Energy-Guided Transport, a novel training-free framework that formulates non-linear physical constraints as a smooth potential energy landscape. Rather than interrupting the ODE integration, our method applies a continuous dissipative force directly to the intermediate generative states, steering them toward the physical manifold. By evaluating this force implicitly via reverse-mode automatic differentiation, our framework preserves the natural dynamics of the flow while maintaining an efficient 𝒪(N) algorithmic complexity. We extensively evaluate our framework across various 1D and 2D physical systems, including the chaotic Navier-Stokes equations. Our method achieves stable generation without structural divergence, significantly improves the stability-accuracy trade-off compared to state-of-the-art methods, and substantially accelerates inference with respect to exact projection strategies. Code available at: https://github.com/ispamm/Projection-free-energy-guided-flow-matching.

Projection-based segregated, iterative, and multi-stage physics-informed neural network for incompressible Navier-stokes equations
Zhenyu Zhang, Zhixiang Ling, Yinghao Huo +2
arXiv5 Nov 2026PreprintarXiv licenceFull textComputational physics Fluid dynamicsarXiv
Hard Constraint Projection in a Physics Informed Neural Network (Burgers)
Miranda J S Horne, Peter K Jimack, Amirul Khan +1
arXiv30 Oct 2026PreprintarXiv licenceFull textFluid dynamicsarXiv
Physics-Informed Neural Networks for coupled stiff transport systems
Laetitia Laguzet, Gabriel Turinici
arXiv26 Oct 2026PreprintCC BY-NC-ND 4.0Full textNumerical analysis & scientific computing Optimisation & controlarXiv
Physics-Informed and Knowledge-Driven Generative AI for Autonomous Discovery of Porous Oxide Energy Materials: Opportunities and Challenges
Dibakar Datta
arXiv2 Nov 2026PreprintCC BY 4.0Full textCondensed matterarXiv
Physics-Informed Neural Networks to Infer the Perpendicular Energy Conductivity in the Scrape-Off Layer of Stellarator Devices
J. Gallego, P. Protopapas, A. Bustos +6
arXiv10 Sep 2026PreprintarXiv licenceFull textPlasma physicsarXiv

Cite

BibTeX RIS

BibTeX

@misc{galadini2026energyguided,
  title = {{Energy-Guided Transport for Projection-Free Physics-Informed Flow Matching}},
  author = {Giuliano Galadini and Aurelio Uncini and Danilo Comminiello},
  howpublished = {arXiv},
  year = {2026},
  month = nov,
  doi = {10.48550/arxiv.2611.06711},
  eprint = {2611.06711},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2611.06711},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Energy-Guided Transport for Projection-Free Physics-Informed Flow Matching
AU  - Giuliano Galadini
AU  - Aurelio Uncini
AU  - Danilo Comminiello
PY  - 2026
DA  - 2026/11/04
PB  - arXiv
DO  - 10.48550/arxiv.2611.06711
UR  - https://arxiv.org/abs/2611.06711
AB  - Deep generative models, particularly Flow Matching frameworks, have emerged as powerful tools for synthesizing complex physical systems. However, standard architectures remain purely statistically driven and agnostic to the underlying governing equations, often violating fundamental physical constraints during inference. Existing zero-shot approaches designed to enforce these constraints typically rely on rigid algebraic projections. While effective for sparse boundary conditions, applying such discrete geometric corrections to dense, full-domain PDEs may disrupt the probability path, \rev{potentially deviating from the learned manifold and introducing substantial computational overhead}. To address these challenges, we introduce Energy-Guided Transport, a novel training-free framework that formulates non-linear physical constraints as a smooth potential energy landscape. Rather than interrupting the ODE integration, our method applies a continuous dissipative force directly to the intermediate generative states, steering them toward the physical manifold. By evaluating this force implicitly via reverse-mode automatic differentiation, our framework preserves the natural dynamics of the flow while maintaining an efficient 𝒪(N) algorithmic complexity. We extensively evaluate our framework across various 1D and 2D physical systems, including the chaotic Navier-Stokes equations. Our method achieves stable generation without structural divergence, significantly improves the stability-accuracy trade-off compared to state-of-the-art methods, and substantially accelerates inference with respect to exact projection strategies. Code available at: https://github.com/ispamm/Projection-free-energy-guided-flow-matching.
ER  -

Abstract

Related

Cite