Metraipolis › archive › dc_48720eee1934

Monotone Viability Learning for Offline Safe Reinforcement Learning — arXiv:2611.07616

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.07616
License (as archived)
CC0-1.0
Archived
2026-11-06T14:50:36.505Z
SHA-256 of the archived text
61699c4c029d95fef074dfc3b0ac70b565c113aff733c6d8ddcc68a7fa433e1b
Stable URL of this page
https://metraipolis.com/documento/dc_48720eee1934

How to cite

Monotone Viability Learning for Offline Safe Reinforcement Learning — arXiv:2611.07616 — arXiv (en), https://arxiv.org/abs/2611.07616 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_48720eee1934 (Metraipolis archive dc_48720eee1934, sha256 61699c4c029d95fef074dfc3b0ac70b565c113aff733c6d8ddcc68a7fa433e1b)
BibTeX
@misc{dc48720eee1934,
  title = {Monotone Viability Learning for Offline Safe Reinforcement Learning — arXiv:2611.07616},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.07616},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_48720eee1934; sha256 61699c4c029d95fef074dfc3b0ac70b565c113aff733c6d8ddcc68a7fa433e1b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.07616. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Offline safe reinforcement learning requires reliable long-horizon safety estimates from fixed data. Learning Hamilton-Jacobi (HJ) viability in this setting calls for semantic preservation, convergence, and numerical stability. To address these challenges, we propose Monotone Viability Learning (MoViL), which learns viable sets from offline transitions through monotone finite-horizon set recursion. The recursion preserves finite-horizon safety semantics and converges to the infinite-horizon viability kernel. Binary supervision maintains a fixed target scale across horizons, avoiding discount-induced numerical attenuation. We further establish sufficient conditions for recovering behavior-supported viable subsets. The learned viability estimates are used to certify actions at deployment. Dubins car experiments validate the numerical accuracy of the learned viable sets. Evaluations on safe RL benchmarks demonstrate state-of-the-art reward-cost performance. MoViL satisfies the average cost constraint on every task and achieves the lowest average cost among the compared methods. These results highlight MoViL's potential to extend long-horizon HJ viability analysis to safety-critical systems where reliable dynamics models are unavailable and online exploration is unsafe.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.07616 — 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 2026-11-06T14:21:42.628Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).