EpicWorldModel: Exploration-driven Planning with Latent World Models

Excerpt

arXiv:2610.05996v1 Announce Type: cross Abstract: Latent world models based on Joint-Embedding Predictive Architecture (JEPA) are deterministic by design. While successful in fully observable scenarios, this paradigm breaks down when past observations and actions lead to multiple plausible future possibilities, e.g., due to occlusion. We introduce EpicWorldModel, a framework to train stochastic JEPAs for environments and tasks with inherent uncertainty under partially observability. We jointly t