{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/verifiably-safe-off-model-reinforcement","title":"Verifiably Safe Off-Model Reinforcement Learning","arxiv_id":"1902.05632","date":"2019-02-14","proceeding":null,"authors":["Nathan Fulton","Andre Platzer"],"abstract":"The desire to use reinforcement learning in safety-critical settings has\ninspired a recent interest in formal methods for learning algorithms. Existing\nformal methods for learning and optimization primarily consider the problem of\nconstrained learning or constrained optimization. Given a single correct model\nand associated safety constraint, these approaches guarantee efficient learning\nwhile provably avoiding behaviors outside the safety constraint. Acting well\ngiven an accurate environmental model is an important pre-requisite for safe\nlearning, but is ultimately insufficient for systems that operate in complex\nheterogeneous environments. This paper introduces verification-preserving model\nupdates, the first approach toward obtaining formal safety guarantees for\nreinforcement learning in settings where multiple environmental models must be\ntaken into account. Through a combination of design-time model updates and\nruntime model falsification, we provide a first approach toward obtaining\nformal safety proofs for autonomous systems acting in heterogeneous\nenvironments.","url_abs":"http://arxiv.org/abs/1902.05632v1","url_pdf":"http://arxiv.org/pdf/1902.05632v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"verifiably-safe-off-model-reinforcement","repo_url":"https://github.com/IBM/vsrl-framework","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"model","task_name":"model"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}