Intelligent agents must be generalists, capable of quickly adapting to various tasks.

Intelligent agents must be generalists, capable of quickly adapting to various tasks. In reinforcement learning (RL), model-based RL learns a dynamics model of the world, in principle enabling transfer to arbitrary reward functions through planning. However, autoregressive model rollouts suffer from compounding error, making model-based RL ineffective for long-horizon problems. Successor features offer an alternative by modeling a policy's long-term state occupancy, reducing policy evaluation under new rewards to linear regression. Yet, policy optimization with successor features can be challenging. This work proposes a novel class of models, i.e., Distributional Successor Features for Zero-Shot Policy Optimization (DiSPOs), that learn a distribution of successor features of a stationary dataset's behavior policy, along with a policy that acts to realize different successor features within the dataset. By directly modeling long-term outcomes in the dataset, DiSPOs avoid compounding error while enabling a simple scheme for zero-shot policy optimization across reward functions. We present a practical instantiation of DiSPOs using diffusion models and show their efficacy as a new class of transferable models, both theoretically and empirically across various simulated robotics problems.

Intelligent agents must be generalists, capable of quickly adapting to various tasks.

Intelligent agents must be generalists, capable of quickly adapting to various tasks. In reinforcement learning (RL), model-based RL learns a dynamics model of the world, in principle enabling transfer to arbitrary reward functions through planning. However, autoregressive model rollouts suffer from compounding error, making model-based RL ineffective for long-horizon problems. Successor features offer an alternative by modeling a policy's long-term state occupancy, reducing policy evaluation under new rewards to linear regression. Yet, policy optimization with successor features can be challenging.

Abstract-- We aim to solve the problem of generating coarseto-fine skills learning from demonstrations (LfD). To scale precision, traditional LfD approaches often rely on extensive fine-grained demonstrations with external interpolations or dynamics models with limited generalization capabilities. For memory-efficient learning and convenient granularity change, we propose a novel diffusion-SSM based policy (DiSPo) that learns from diverse coarse skills and produces varying control scales of actions by leveraging a state-space model, Mamba. Our evaluations show the adoption of Mamba and the proposed step-scaling method enables DiSPo to outperform in five coarseto-fine benchmark tests while DiSPo shows decent performance in typical fine-grained motion learning and reproduction. We finally demonstrate the scalability of actions with simulation and real-world manipulation tasks. In typical object manipulation, small imprecision around local regions often leads to the failure of entire tasks, such Figure 1: A capture of a square-drawing task that requires as robot welding, screwing, and drawing, as shown in Figure 1.