← back to paper
arxiv: 2604.26694 · 2 revisions
Unified 4D World Action Modeling from Video Priors with Asynchronous Denoising