MODULI: Unlocking Preference Generalization via Diffusion Models for Offline Multi-Objective Reinforcement Learning

no code implementations • 28 Aug 2024 • Yifu Yuan, Zhenrui Zheng, Zibin Dong, Jianye Hao

Leveraging the excellent expressive and generalization capabilities of diffusion models, we propose MODULI (Multi-objective Diffusion Planner with Sliding Guidance), which employs a preference-conditioned diffusion model as a planner to generate trajectories that align with various preferences and derive action for decision-making.

Multi-Objective Reinforcement Learning