English

Ratio-Variance Regularized Policy Optimization

Machine Learning 2026-05-27 v1 Artificial Intelligence

Abstract

Standard on-policy reinforcement learning relies on heuristic clipping to enforce trust regions, but this mechanism imposes a severe cost by indiscriminately truncating high-return yet high-divergence updates. We demonstrate that explicitly constraining the policy ratio variance provides a principled local approximation to trust-region constraints, eliminating the need for binary hard clipping. By acting as a distributional ``soft brake'', this approach preserves critical gradient signals from novel discoveries while naturally down-weighting and enabling the reuse of stale, off-policy data. We introduce R2VPO{\bf R}^2{\bf VPO} (Ratio-Variance Regularized Policy Optimization), which implements this constraint via a primal-dual optimization framework. Extensive evaluations across 77 LLM scales, spanning both fast and slow reasoning paradigms, and 1010 robotic control tasks demonstrate the generality of the proposed approach. R2^2VPO achieves substantial performance gains on mathematical reasoning benchmarks, with particularly pronounced improvements on smaller models, while significantly improving sample efficiency. Furthermore, it consistently outperforms PPO baselines in continuous control domains, particularly in sparse-reward and dynamic environments. Together, these findings establish ratio-variance regularization as a principled foundation for stable and data-efficient policy optimization.

Keywords

Cite

@article{arxiv.2605.26784,
  title  = {Ratio-Variance Regularized Policy Optimization},
  author = {Yu Luo and Shuo Han and Yihan Hu and Lei Lv and Huaping Liu and Fuchun Sun and Jianye Hao and Dong Li},
  journal= {arXiv preprint arXiv:2605.26784},
  year   = {2026}
}