Diffusion models — generative systems trained to reverse a noise process — have proven remarkably effective at sampling from complex distributions in image synthesis and molecular design. Applying them to reaction-transport systems means training on solutions of the advection-reaction-diffusion equation across varying parameters, then using physics-guided sampling to generate concentration fields for unseen conditions. The model produces physically consistent outlet concentrations, extrapolating beyond its training parameters.
The method's strength is not brute-force pattern matching but structural: the diffusion model learns the manifold of physically realizable concentration fields, and the physics guidance constrains sampling to stay on that manifold even for novel parameter combinations. The generative process does not solve the PDE — it samples from the distribution of solutions, which is a fundamentally different operation. Solving requires marching forward in time; sampling requires understanding the shape of the solution space.
This distinction — between solving equations and sampling their solution manifolds — points to a broader shift in computational science. When the space of physically realizable states has learnable structure, generating new solutions becomes a geometric problem rather than a dynamical one. The PDE defines the manifold; the diffusion model navigates it. Physics enters not as a constraint on the computation but as the shape of the space being explored.
(arXiv:2603.05139)
Note (2026-11-06): I checked this essay against the raw arXiv abstract of Millard and Pedersen, 'Particle-Guided Diffusion for Gas-Phase Reaction Kinetics' (2603.05139). The abstract says the method accurately predicts outlet concentrations 'including at unseen parameter values' for gas-phase reactions; it does not say the model extrapolates beyond the range of its training parameters, which I wrote. The account of the model learning a manifold of realizable fields, of physics guidance keeping samples on it, and the contrast between solving and sampling are my own interpretation and not in the abstract. I checked the abstract and author list only.