# Gradient-Based Planning for Long-Horizon World Models GRASP makes long-horizon planning practical with learned dynamics. - **Source:** [Berkeley AI Research Blog](https://bair.berkeley.edu/blog) - **Published:** 2026-04-20 - **Topic:** [Data & AI](https://brifing.si/c/data-ai) - **Original article:** - **On Brifing:** https://brifing.si/s/e6ff6801-a9db-4131-a10c-b8c180b0d0fd ## What is GRASP? GRASP is a gradient-based planner for learned dynamics that makes long-horizon planning practical by lifting trajectories into virtual states, adding stochasticity, and reshaping gradients. ## World Models World models are learned models that predict future outcomes given the current state and future actions. These models can predict long sequences in high-dimensional visual spaces. ## Planning Process Planning involves optimizing action sequences through the model to minimize terminal error, given a start state and a goal. ## Challenges of Long-Horizon Planning Long-horizon planning is challenging due to deep and ill-conditioned computation graphs, landscapes full of traps, and issues specific to learned models. ## GRASP's Solutions GRASP addresses long-horizon planning challenges by lifting into virtual states, adding stochasticity, and reshaping gradients for robustness. ## Lifting into Virtual States GRASP lifts trajectories into virtual states to ensure optimization is parallel across time. ## Adding Stochasticity GRASP adds stochasticity directly to state iterates for exploration. ## Reshaping Gradients GRASP reshapes gradients to ensure actions receive clean signals while avoiding brittle 'state-input' gradients through high-dimensional vision models. ## Application Areas GRASP can be used in robotics, autonomous vehicles, and other fields requiring long-horizon planning. ## Future Potential GRASP has the potential to expand the use of learned world models by making long-horizon planning more practical and robust. ## Conclusion GRASP provides an effective solution to the challenges of long-horizon planning and holds significant potential for future applications. ## Sources covering this story - http://bair.berkeley.edu/blog/2026/04/20/grasp/ — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._