Search papers, labs, and topics across Lattice.
This paper introduces Bidirectional Conditional Flow Matching (Bi-CFM) to address the challenging inverse problems in chaotic systems, specifically inferring initial conditions from final states. By learning bidirectional mappings between initial and final state distributions, Bi-CFM effectively captures the stochastic nature of chaotic evolution and reduces error accumulation, achieving significant improvements in distribution-level metrics. The method also extends to Conservation-constrained Bi-CFM (CBi-CFM), which maintains conservation laws in chaotic systems, demonstrating superior accuracy and speed in applications ranging from classic dynamical systems to real-world observations of globular clusters.
Bi-CFM achieves over two orders of magnitude speedup while improving accuracy in solving inverse problems of chaotic systems, even under conservation constraints.
Modeling chaotic systems is crucial yet challenging. Inverse problems in chaotic dynamics, namely inferring initial conditions from final states, remain largely unsolved because of ill-posedness, non-uniqueness, instability, and potentially chaotic time-reverse dynamics. We address this open problem with Bidirectional Conditional Flow Matching (Bi-CFM), which learns bidirectional mappings between distributions of initial and final states to capture the stochasticity of chaotic evolution and mitigate exponential error accumulation over time. Furthermore, for systems with conservation laws, we extend it to Conservation-constrained Bi-CFM (CBi-CFM). Across the classic Lorenz, Circuit, and high-dimensional Lorenz 96 systems, Bi-CFM improves five distribution-level metrics over baselines while achieving a speedup of more than two orders of magnitude. In the three-body planet-planet scattering problem in planetary dynamics, CBi-CFM better respects conservation laws, with conservation errors comparable to those of the ground truth. Finally, on real observations of globular clusters, collisional million-body systems shaped by $\sim 10^{10}$ years (10 Gyr) of evolution, our method represents an advance in accuracy, establishing a scalable route to solving inverse problems of long-timescale real-world chaotic dynamics.
