This is the pre-proceedings for the RLC 2026. You may expect minor changes.
Reinforcement Learning Journal, vol. 7, 2026, pp. TBD.
Will be presented at the Reinforcement Learning Conference (RLC), Montréal, Quebec, Canada, August 15–17, 2026.
Keywords: Safe Reinforcement Learning, Skill-based Reinforcement Learning, Risk Planning.
Safe Reinforcement Learning (Safe RL) aims to ensure safety when an RL agent conducts learning by interacting with real-world environments where improper actions can induce high costs or lead to severe consequences. In this paper, we propose a novel Safe Skill Planning (SSkP) approach to enhance effective safe RL by exploiting auxiliary offline demonstration data. SSkP involves a two-stage process. First, we employ Positive-Unlabeled (PU) learning to learn a skill risk predictor from the offline demonstration data. Then, based on the learned skill risk predictor, we develop a novel risk planning process to enhance online safe RL and learn a risk-averse safe policy efficiently through interactions with the online RL environment, while simultaneously adapting the skill risk predictor to the environment. We conduct experiments in several robotic simulation environments. The experimental results demonstrate that the proposed approach consistently outperforms previous state-of-the-art safe RL methods.
Hanping Zhang and Yuhong Guo. "Skill-based Safe Reinforcement Learning with Risk Planning." Reinforcement Learning Journal, vol. 7, 2026, pp. TBD.
BibTeX:@article{zhang2026skillbased,
title={Skill-based Safe Reinforcement Learning with Risk Planning},
author={Hanping Zhang and Yuhong Guo},
journal={Reinforcement Learning Journal},
volume={7},
pages={},
year={2026}
}