Back to the Future: Reverse-Consequence Modeling for Long-Horizon Particle Dynamics

arXiv:2610.18168v1 Announce Type: new Abstract: Accurate prediction of object trajectories is fundamental to mechanical understanding and robotic control. However, autoregressive models inherently suffer from cumulative errors, leading to severe drift in long-horizon rollouts. Observing that in many physical systems with dissipative or convergence-to-rest behavior, the final states are often more predictable than transient dynamics (e.g., a sliding object resting under friction or a pendulum stabilizing at equilibrium), we propose ReStep, a cascaded approach that reverses the conventional forward-prediction paradigm. ReStep first employs a stochastic model to directly predict the final state, then leverages Anchored Diffusion to generate goal-conditioned intermediate trajectories. By conditioning on both ''boundary states'', it effectively constrains dynamics and reduces uncertainty, producing physically plausible and temporally coherent transitions. We further integrate ReStep dynamics as the physical engine within a 3D Gaussian Splatting pipeline, enabling a unified physical-visual framework for intuitive inference from images. Extensive experiments on diverse rigid and non-rigid collision datasets show that ReStep achieves superior long-term accuracy, substantially mitigating error accumulation via reverse-consequence modeling.

Related

• Parallel Predictive World Models for Accurate and Efficient Long-Horizon Planning
• From Fixed Grids to Moving Particles:A Transferable Latent Operator for Fluid Dynamics
• ACID: Action Consistency via Inverse Dynamics for Planning with World Models
• PDE-regularized Dynamics-informed Diffusion with Uncertainty-aware Filtering for Long-Horizon Dynamics
• When Can World Models Recover Physical Laws?

Source: arXiv cs.AI | 2026-10-15