3 papers
cs.LG2025
Enforcing governing equation constraints in neural PDE solvers via training-free projections
Omer Rochman, Gilles Louppe
Neural PDE solvers used for scientific simulation often violate governing equation constraints. While linear constraints can be projected cheaply, many constraints are nonlinear, c…
cs.LG2025
Training-Free Data Assimilation with GenCast
Thomas Savary, François Rozet, Gilles Louppe
Data assimilation is widely used in many disciplines such as meteorology, oceanography, and robotics to estimate the state of a dynamical system from noisy observations. In this wo…
cs.LG2025
Lost in Latent Space: An Empirical Study of Latent Diffusion Models for Physics Emulation
François Rozet, Ruben Ohana, Michael McCabe +3
The steep computational cost of diffusion models at inference hinders their use as fast physics emulators. In the context of image and video generation, this computational drawback…