Search papers, labs, and topics across Lattice.
This paper introduces a deep learning framework called PAPS that efficiently solves transient Fokker-Planck equations (FPEs) for arbitrary parameters and initial distributions by unifying initial, transient, and stationary distributions via Gaussian mixture distributions (GMDs). A constraint-preserving autoencoder maps constrained GMD parameters to unconstrained latent representations, enabling a single evolution network to model transient dynamics. Experiments show PAPS achieves inference speeds four orders of magnitude faster than GPU-accelerated Monte Carlo simulations while maintaining high accuracy.
Unlock real-time parameter sweeps and systematic investigations of stochastic bifurcations with a deep learning framework that solves Fokker-Planck equations 10,000x faster than Monte Carlo.
Efficiently solving the Fokker-Planck equation (FPE) is central to analyzing complex parameterized stochastic systems. However, current numerical methods lack parallel computation capabilities across varying conditions, severely limiting comprehensive parameter exploration and transient analysis. This paper introduces a deep learning-based pseudo-analytical probability solution (PAPS) that, via a single training process, simultaneously resolves transient FPE solutions for arbitrary multi-modal initial distributions, system parameters, and time points. The core idea is to unify initial, transient, and stationary distributions via Gaussian mixture distributions (GMDs) and develop a constraint-preserving autoencoder that bijectively maps constrained GMD parameters to unconstrained, low-dimensional latent representations. In this representation space, the panoramic transient dynamics across varying initial conditions and system parameters can be modeled by a single evolution network. Extensive experiments on paradigmatic systems demonstrate that the proposed PAPS maintains high accuracy while achieving inference speeds four orders of magnitude faster than GPU-accelerated Monte Carlo simulations. This efficiency leap enables previously intractable real-time parameter sweeps and systematic investigations of stochastic bifurcations. By decoupling representation learning from physics-informed transient dynamics, our work establishes a scalable paradigm for probabilistic modeling of multi-dimensional, parameterized stochastic systems.
