--- title: Gradient-based Planning for World Models at Longer Horizons description: GRASP is a new gradient-based planner for learned world models that makes long-horizon planning practical. It addresses three core problems: (1) it lifts the... canonical: https://daily.dev/posts/gradient-based-planning-for-world-models-at-longer-horizons-mfy611yao twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Gradient-based Planning for World Models at Longer Horizons | daily.dev og:description: GRASP is a new gradient-based planner for learned world models that makes long-horizon planning practical. It addresses three core problems: (1) it lifts the... og:url: https://daily.dev/posts/gradient-based-planning-for-world-models-at-longer-horizons-mfy611yao og:image: https://api.daily.dev/og/posts/mfY611YAo.png og:image:alt: Gradient-based Planning for World Models at Longer Horizons og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Gradient-based Planning for World Models at Longer Horizons **[BAIR](https://daily.dev/sources/bair)** · 14 min read · 0 upvotes · 0 comments ## Summary GRASP is a new gradient-based planner for learned world models that makes long-horizon planning practical. It addresses three core problems: (1) it lifts the trajectory into virtual states enabling parallel optimization across time, (2) injects Gaussian noise into state iterates for exploration, and (3) stops brittle state-input gradients through the world model while preserving clean action gradients. The key insight is that state Jacobians in deep learning-based world models suffer from adversarial robustness issues (similar to adversarial examples in vision), making direct state optimization unreliable, while action Jacobians remain well-behaved. GRASP combines a stop-gradient collocation objective with dense goal shaping and periodic synchronization with true rollout gradients. Benchmarks on Push-T tasks show GRASP outperforms CEM, vanilla gradient descent, and LatCo at horizons H=50–80, achieving higher success rates and faster planning times. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [NVIDIA Research Unlocks Advanced Grasping, Smarter Autonomous Driving and Agent Training at Scale](https://daily.dev/posts/nvidia-research-unlocks-advanced-grasping-smarter-autonomous-driving-and-agent-training-at-scale-mvjwdapsk) · NVIDIA · 0 upvotes · 0 comments - [Toward DeepSeek-R1: Building Thinking Models Using GRPO](https://daily.dev/posts/toward-deepseek-r1-building-thinking-models-using-grpo-zm2qycu1b) · PyImageSearch · 0 upvotes · 0 comments - [R²D²: Perception-Guided Task & Motion Planning for Long-Horizon Manipulation](https://daily.dev/posts/r-d-perception-guided-task-motion-planning-for-long-horizon-manipulation-ymtfswai7) · NVIDIA Developer · 0 upvotes · 0 comments - [A General Goal-Conditioned Minecraft Model](https://daily.dev/posts/a-general-goal-conditioned-minecraft-model-ydhsvavpp) · Hacker News · 0 upvotes · 0 comments - [How to Beat GRPO Without Touching Model Weights](https://daily.dev/posts/how-to-beat-grpo-without-touching-model-weights-8m9avbzph) · Daily Dose of Data Science \| Avi Chawla \| Substack · 14 upvotes · 0 comments --- Tags: [#reinforcement-learning](https://daily.dev/tags/reinforcement-learning), [#world-models](https://daily.dev/tags/world-models) [View this post on daily.dev](https://daily.dev/posts/gradient-based-planning-for-world-models-at-longer-horizons-mfy611yao) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Gradient-based Planning for World Models at Longer Horizons","url":"https://daily.dev/posts/gradient-based-planning-for-world-models-at-longer-horizons-mfy611yao","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/gradient-based-planning-for-world-models-at-longer-horizons-mfy611yao"},"datePublished":"2026-04-20T21:43:21.563Z","dateModified":"2026-10-14T02:33:41.042Z","description":"GRASP is a new gradient-based planner for learned world models that makes long-horizon planning practical. It addresses three core problems: (1) it lifts the...","image":"https://media.daily.dev/image/fetch/s--Jpfllqoz--/f_auto,pg_1,q_auto/https://bair.berkeley.edu/static/blog/grasp/manifold_adversarial.gif?_a=BAMAMiZY0","thumbnailUrl":"https://media.daily.dev/image/fetch/s--Jpfllqoz--/f_auto,pg_1,q_auto/https://bair.berkeley.edu/static/blog/grasp/manifold_adversarial.gif?_a=BAMAMiZY0","isAccessibleForFree":true,"articleSection":"BAIR","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"BAIR","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/bair","url":"https://daily.dev/sources/bair"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/gradient-based-planning-for-world-models-at-longer-horizons-mfy611yao","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"reinforcement-learning,world-models","timeRequired":"PT14M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"BAIR","item":"https://daily.dev/sources/bair"},{"@type":"ListItem","position":3,"name":"Gradient-based Planning for World Models at Longer Horizons"}]} ```