Latent Generative Solvers for Generalizable Long-Term Physics Simulation

• URL: http://arxiv.org/abs/2602.11229v1
• Date: Wed, 11 Feb 2026 15:34:52 GMT
• Title: Latent Generative Solvers for Generalizable Long-Term Physics Simulation
• Authors: Zituo Chen, Haixu Wu, Sili Deng,
• Abstract summary: Latent Generative Solvers (LGS) is a framework that maps diverse PDE states into a shared latent physics space with a pretrained VAE.<n>Our key mechanism is an uncertainty knob that perturbs latent inputs during training and inference, teaching the solver to correct off-manifold rollout drift.<n>LGS matches strong deterministic neural-operator baselines on short horizons while substantially reducing rollout drift on long horizons.
• Score: 12.894423121609526
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We study long-horizon surrogate simulation across heterogeneous PDE systems. We introduce Latent Generative Solvers (LGS), a two-stage framework that (i) maps diverse PDE states into a shared latent physics space with a pretrained VAE, and (ii) learns probabilistic latent dynamics with a Transformer trained by flow matching. Our key mechanism is an uncertainty knob that perturbs latent inputs during training and inference, teaching the solver to correct off-manifold rollout drift and stabilizing autoregressive prediction. We further use flow forcing to update a system descriptor (context) from model-generated trajectories, aligning train/test conditioning and improving long-term stability. We pretrain on a curated corpus of $\sim$2.5M trajectories at $128^2$ resolution spanning 12 PDE families. LGS matches strong deterministic neural-operator baselines on short horizons while substantially reducing rollout drift on long horizons. Learning in latent space plus efficient architectural choices yields up to \textbf{70$\times$} lower FLOPs than non-generative baselines, enabling scalable pretraining. We also show efficient adaptation to an out-of-distribution $256^2$ Kolmogorov flow dataset under limited finetuning budgets. Overall, LGS provides a practical route toward generalizable, uncertainty-aware neural PDE solvers that are more reliable for long-term forecasting and downstream scientific workflows.

Related papers

• Operator Learning Using Weak Supervision from Walk-on-Spheres [81.26322147849918]
  Training neural PDE solvers is often bottlenecked by expensive data generation or unstable physics-informed neural network (PINN)<n>We propose an alternative approach using Monte Carlo approaches to estimate the solution to the PDE as a process for weak supervision during training.
  arXiv  Detail & Related papers  (2026-03-01T17:23:39Z)
• ResAD: Normalized Residual Trajectory Modeling for End-to-End Autonomous Driving [64.42138266293202]
  ResAD is a Normalized Residual Trajectory Modeling framework.<n>It reframes the learning task to predict the residual deviation from an inertial reference.<n>On the NAVSIM benchmark, ResAD achieves a state-of-the-art PDMS of 88.6 using a vanilla diffusion policy.
  arXiv  Detail & Related papers  (2025-10-09T17:59:36Z)
• Flow marching for a generative PDE foundation model [0.0]
  We propose Flow Marching, an algorithm that bridges neural operator learning with flow matching motivated by an analysis of error accumulation in physical dynamical systems.<n>We also introduce a Physics-Pretrained Variational Autoencoder (P2E) to embed physical trajectories into a compact latent space.<n>We curate a corpus of 2.5M trajectories across 12 distinct PDE families and train suites of P2Es and FMTs at multiple scales.
  arXiv  Detail & Related papers  (2025-09-23T04:00:41Z)
• Physics-informed Reduced Order Modeling of Time-dependent PDEs via Differentiable Solvers [1.224954637705144]
  We propose Physics-informed ROM ($Phi$-ROM) by incorporating differentiable PDE solvers into the training procedure.<n>Our model outperforms state-of-the-art data-driven ROMs and other physics-informed strategies.<n>$Phi$-ROM learns to recover and forecast the solution fields even when trained or evaluated with sparse and irregular observations of the fields.
  arXiv  Detail & Related papers  (2025-05-20T16:47:04Z)
• Implicit Neural Differential Model for Spatiotemporal Dynamics [5.1854032131971195]
  We introduce Im-PiNDiff, a novel implicit physics-integrated neural differentiable solver for stabletemporal dynamics.<n>Inspired by deep equilibrium models, Im-PiNDiff advances the state using implicit fixed-point layers, enabling robust long-term simulation.<n>Im-PiNDiff achieves superior predictive performance, enhanced numerical stability, and substantial reductions in memory and cost.
  arXiv  Detail & Related papers  (2025-04-03T04:07:18Z)
• DPOT: Auto-Regressive Denoising Operator Transformer for Large-Scale PDE Pre-Training [87.90342423839876]
  We present a new auto-regressive denoising pre-training strategy, which allows for more stable and efficient pre-training on PDE data. We train our PDE foundation model with up to 0.5B parameters on 10+ PDE datasets with more than 100k trajectories.
  arXiv  Detail & Related papers  (2024-03-06T08:38:34Z)
• Koopman Invertible Autoencoder: Leveraging Forward and Backward Dynamics for Temporal Modeling [13.38194491846739]
  We propose a novel machine learning model based on Koopman operator theory, which we call Koopman Invertible Autoencoders (KIA) KIA captures the inherent characteristic of the system by modeling both forward and backward dynamics in the infinite-dimensional Hilbert space. This enables us to efficiently learn low-dimensional representations, resulting in more accurate predictions of long-term system behavior.
  arXiv  Detail & Related papers  (2023-09-19T03:42:55Z)
• Implicit Stochastic Gradient Descent for Training Physics-informed Neural Networks [51.92362217307946]
  Physics-informed neural networks (PINNs) have effectively been demonstrated in solving forward and inverse differential equation problems. PINNs are trapped in training failures when the target functions to be approximated exhibit high-frequency or multi-scale features. In this paper, we propose to employ implicit gradient descent (ISGD) method to train PINNs for improving the stability of training process.
  arXiv  Detail & Related papers  (2023-03-03T08:17:47Z)
• MAgNet: Mesh Agnostic Neural PDE Solver [68.8204255655161]
  Climate predictions require fine-temporal resolutions to resolve all turbulent scales in the fluid simulations. Current numerical model solveers PDEs on grids that are too coarse (3km to 200km on each side) We design a novel architecture that predicts the spatially continuous solution of a PDE given a spatial position query.
  arXiv  Detail & Related papers  (2022-10-11T14:52:20Z)
• Learning to Accelerate Partial Differential Equations via Latent Global Evolution [64.72624347511498]
  Latent Evolution of PDEs (LE-PDE) is a simple, fast and scalable method to accelerate the simulation and inverse optimization of PDEs. We introduce new learning objectives to effectively learn such latent dynamics to ensure long-term stability. We demonstrate up to 128x reduction in the dimensions to update, and up to 15x improvement in speed, while achieving competitive accuracy.
  arXiv  Detail & Related papers  (2022-06-15T17:31:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.