Proceedings

A Case for Validation Buffer in Pessimistic Actor-Critic

Michał Nauman, Mateusz Ostaszewski, Marek Cygan

Download BibTeX Cite

Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence. Main Track. Pages 5976-5984.

https://doi.org/10.24963/ijcai.2025/665

Keywords

  • Robotics: ROB: Behavior and control
  • Planning and Scheduling: PS: Markov decisions processes
  • Machine Learning: ML: Reinforcement learning
  • Robotics: ROB: Learning in robotics

Abstract

In this paper, we investigate the issue of error accumulation in critic networks updated via pessimistic temporal difference objectives. We show that the critic approximation error can be approximated via a recursive fixed-point model similar to that of the Bellman value. We use such recursive definition to retrieve the conditions under which the pessimistic critic is unbiased. Building on these insights, we propose Validation Pessimism Learning (VPL) algorithm. VPL uses a small validation buffer to adjust the levels of pessimism throughout the agent training, with the pessimism set such that the approximation error of the critic targets is minimized. We investigate the proposed approach on a variety of locomotion and manipulation tasks and report improvements in sample efficiency and performance.

Michał Nauman, Mateusz Ostaszewski, and Marek Cygan. 2025. A Case for Validation Buffer in Pessimistic Actor-Critic. In Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25. IJCAI, 5976–5984. https://doi.org/10.24963/ijcai.2025/665
Welcome Future Conferences AI Journal Division IJCAI Proceedings About IJCAI IJCAI Awards