Unlocking Temporal Generalization in Hamiltonian Video Dynamics Models

Eli Laird, Corey Clark

World models are typically trained to predict discrete-time physical dynamics with a fixed step size baked into the model weights, preventing prediction at variable temporal resolu…