Research

JEPA world models + Hierarchical Planning is a massive step for long-horizon robotics. A classic failure mode I’ve faced with planning with …

JEPA world models + Hierarchical Planning is a massive step for long-horizon robotics. A classic failure mode I’ve faced with planning with world models: flat planning often 'cheats.' For example, in

DGX agentx-post
researchyann-lecun--x

JEPA world models + Hierarchical Planning is a massive step for long-horizon robotics. A classic failure mode I’ve faced with planning with world models: flat planning often "cheats." For example, in pick-and-place, the robot often reaches the target state in imagination without actually picking the object. Hierarchical layers fix this by first optimizing for valid subgoals (like the grip) before the finish. Incredible work, @kevinghstz and team! Huge congratulations. 🚀 Hierarchical planning unlocks long-horizon, non-greedy behavior in JEPA world models. Paper: https://arxiv.org/pdf/2604.03208 Website: https://kevinghst.github.io/HWM/ Code: https://github.com/kevinghst/HWM_PLDM

Related

Source: Yann LeCun (X) | 2026-04-08

Loading related sources…