Metraipolis › archive › dc_7d101e7e14f2
A Unified and Constrained View of Regularization-Based Robust Reinforcement Learning — arXiv:2609.13050
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.13050
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T03:15:56.764Z
- SHA-256 of the archived text
a9b48f6dbaae198c670f99613d99b9f20b87c7d2fdb594ba1a7e4f58dba8e980- Stable URL of this page
https://metraipolis.com/documento/dc_7d101e7e14f2
How to cite
A Unified and Constrained View of Regularization-Based Robust Reinforcement Learning — arXiv:2609.13050 — arXiv (en), https://arxiv.org/abs/2609.13050 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7d101e7e14f2 (Metraipolis archive dc_7d101e7e14f2, sha256 a9b48f6dbaae198c670f99613d99b9f20b87c7d2fdb594ba1a7e4f58dba8e980)
BibTeX
@misc{dc7d101e7e14f2,
title = {A Unified and Constrained View of Regularization-Based Robust Reinforcement Learning — arXiv:2609.13050},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.13050},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7d101e7e14f2; sha256 a9b48f6dbaae198c670f99613d99b9f20b87c7d2fdb594ba1a7e4f58dba8e980},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.13050. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Regularization-based methods have become a standard approach for training Deep Reinforcement Learning policies against adversarial input perturbations. In this paper, we unify these methods by deriving new upper bounds on the performance gap between the nominal and worst-case policies. Each upper bound is expressed as an existing regularization objective plus a KL-divergence penalty between the nominal and worst-case policies, which further explains why adding a KL penalty improves robustness in practice. Building on these bounds, we formulate robust training as a constrained optimization problem, showing that existing methods correspond to the special case of a fixed Lagrange multiplier. We instead update the multiplier jointly with the policy to automatically tune the regularization weight. Finally, we conduct extensive adversarial evaluations across several continuous control tasks to validate our theoretical analysis. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-14 in cs.LG). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.13050 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-16T03:05:44.085Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).