world model
In reinforcement learning and robotics, a world model is a generative model of the environment that an agent uses for planning and decision-making, enabling the agent to simulate possible future states to optimize actions without direct interaction.
- A Smooth Sea Never Made a Skilled SAILOR: Robust Imitation via Learning to Search
- Bootstrap Off-policy with World Model
- Curious Causality-Seeking Agents Learn Meta Causal World
- Dynamics-Aligned Latent Imagination in Contextual World Models for Zero-Shot Generalization
- Explainably Safe Reinforcement Learning
- FutureSightDrive: Thinking Visually with Spatio-Temporal CoT for Autonomous Driving
- Learning and Planning Multi-Agent Tasks via an MoE-based World Model
- Modelling the control of offline processing with reinforcement learning
- OSVI-WM: One-Shot Visual Imitation for Unseen Tasks using World-Model-Guided Trajectory Generation