Hierarchical World Models with Implicit Dynamics
cs.RO, cs.AI
Submitted: 2026-10-30
Updated: 2026-10-30
Project page: https://id-wm.github.io
License: http://creativecommons.org/licenses/by/4.0/
The gist: Planning with world models enables test-time adaptation for embodied agents, offering superior generalization over direct policy learning.
Terminology
Abstract
Planning with world models enables test-time adaptation for embodied agents, offering superior generalization over direct policy learning. However, long-horizon tasks remain hindered by cumulative errors and search complexity. Existing hierarchical methods often use rigid macro-actions or latent skills, leading to brittle representations and limited fine-grained control. We introduce ID-WM (Implicit Dynamics World Models), a hierarchical world model that decouples state-space reachability from local control. Our framework utilizes a high-level model to sample feasible transitions across the state manifold and a low-level world model as an inverse dynamics solver to ground these transitions into precise actions. This allows the planner to identify reachable subgoals without being constrained by low-level execution details. By prioritizing local physical feasibility over global behavioral patterns, ID-WM synthesizes novel long-horizon plans from short segments, transcending the specific behavioral modes seen during training. We demonstrate its effectiveness across three navigation and manipulation suites, achieving a 38.5% improvement over state-of-the-art policy and diffusion-based planners. Crucially, by leveraging this hierarchical structure to mitigate compounding errors, our approach yields a 135% performance gain over single-level world model MPC, particularly in complex, unseen configurations for long-horizon tasks. Videos can be viewed at https://id-wm.github.io
Sources
- World Models
- GAIA-1: A Generative World Model for Autonomous Driving
- Training Agents Inside of Scalable World Models
- Learning Massively Multitask World Models for Continuous Control
- Causal World Modeling for Robot Control
- V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning
- World Models for Learning Dexterous Hand-Object Interactions from Human Videos
- Planning with Reasoning using Vision Language World Model
- Relay Policy Learning: Solving Long-Horizon Tasks via Imitation and Reinforcement Learning
- CHD: Coupled Hierarchical Diffusion for Long-Horizon Tasks
- Hierarchical Planning with Latent World Models
- Parallel Stochastic Gradient-Based Planning for World Models
- Hierarchical Multiscale Recurrent Neural Networks
- Hierarchical Foresight: Self-Supervised Learning of Long-Horizon Tasks via Visual Subgoal Generation
- Leveraging Jumpy Models for Planning and Fast Learning in Robotic Domains
- Open-Sora 2.0: Training a Commercial-Level Video Generation Model in $200k
- Video models are zero-shot learners and reasoners
- Mastering Diverse Domains through World Models
- Temporal Difference Learning for Model Predictive Control
- Closing the Train-Test Gap in World Models for Gradient-Based Planning
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation