Bidirectional Autoregressive Latent Diffusion for Forward and Inverse Magnetohydrodynamics

2026-06-28

AI Summary

This work addresses the challenge of predicting multi-physics field evolution—encompassing density, pressure, velocity, and magnetic fields—and solving inverse problems in magnetohydrodynamics (MHD) under the scarcity of ground-truth data. To this end, the authors propose a bidirectional autoregressive latent diffusion model, which, for the first time, integrates a bidirectional autoregressive diffusion mechanism into MHD systems. By enforcing consistency between forward and backward temporal evolutions, the model establishes a self-supervised constraint that enables uncertainty quantification at test time without requiring true reference data. Coupled with sparse observation fusion and an adaptive feedback mechanism, the approach achieves high-fidelity multi-field evolution prediction and facilitates non-intrusive plasma diagnostics, demonstrating strong effectiveness and robustness in practical applications.

Abstract

This work presents a new bidirectional autoregressive latent diffusion approach for predicting the evolution of multiple fields (mass density, pressure, velocity, and magnetic field components) for magnetohydrodynamics. We show that this bidirectional flow can be used as a self-supervised consistency metric for uncertainty and error estimation, which enables the model to estimate test-time uncertainty and error without access to ground truth, by comparing how closely flowing forwards and backwards in time returns to the same predicted fields. We also demonstrate this methods's potential to serve as a non-invasive plasma diagnostic, and show how adaptive feedback can be used to make the model more robust based on sparse diagnostics or limited views/measurements.

Problem

• magnetohydrodynamics
• inverse problem
• uncertainty estimation
• plasma diagnostics
• field evolution

Innovation

• bidirectional autoregressive latent diffusion
• self-supervised consistency
• uncertainty estimation
• plasma diagnostics
• adaptive feedback