{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:IBK22QQUL5OZ7RCAKKHJP7DFOE","short_pith_number":"pith:IBK22QQU","schema_version":"1.0","canonical_sha256":"4055ad42145f5d9fc440528e97fc65712a6d80ca6c43107d40eddbd5bafa66f7","source":{"kind":"arxiv","id":"2011.05869","version":3},"attestation_state":"computed","paper":{"title":"CRPO: A New Approach for Safe Reinforcement Learning with Convergence Guarantee","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Guanghui Lan, Tengyu Xu, Yingbin Liang","submitted_at":"2020-11-11T16:05:14Z","abstract_excerpt":"In safe reinforcement learning (SRL) problems, an agent explores the environment to maximize an expected total reward and meanwhile avoids violation of certain constraints on a number of expected total costs. In general, such SRL problems have nonconvex objective functions subject to multiple nonconvex constraints, and hence are very challenging to solve, particularly to provide a globally optimal policy. Many popular SRL algorithms adopt a primal-dual structure which utilizes the updating of dual variables for satisfying the constraints. In contrast, we propose a primal approach, called const"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2011.05869","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2020-11-11T16:05:14Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"1b46dc824d0da08e63ba056b027d713881d6d9379a879f16231bf89644a95c84","abstract_canon_sha256":"544173e5385476ce4d6e6207092fb2bcd5c68fbc52f7db6a1d01386eb7ea1a3c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:44:34.467879Z","signature_b64":"Sd9FsrlHmQDb9fZ6SUPni6nfrP2Gl4kVdsL/nc/8/Qo/JE41OFfhhxRrniS427pI6EkN7GePyPBr/2lAWHUaDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4055ad42145f5d9fc440528e97fc65712a6d80ca6c43107d40eddbd5bafa66f7","last_reissued_at":"2026-07-05T02:44:34.467390Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:44:34.467390Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CRPO: A New Approach for Safe Reinforcement Learning with Convergence Guarantee","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Guanghui Lan, Tengyu Xu, Yingbin Liang","submitted_at":"2020-11-11T16:05:14Z","abstract_excerpt":"In safe reinforcement learning (SRL) problems, an agent explores the environment to maximize an expected total reward and meanwhile avoids violation of certain constraints on a number of expected total costs. In general, such SRL problems have nonconvex objective functions subject to multiple nonconvex constraints, and hence are very challenging to solve, particularly to provide a globally optimal policy. Many popular SRL algorithms adopt a primal-dual structure which utilizes the updating of dual variables for satisfying the constraints. In contrast, we propose a primal approach, called const"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2011.05869","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2011.05869/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2011.05869","created_at":"2026-07-05T02:44:34.467457+00:00"},{"alias_kind":"arxiv_version","alias_value":"2011.05869v3","created_at":"2026-07-05T02:44:34.467457+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2011.05869","created_at":"2026-07-05T02:44:34.467457+00:00"},{"alias_kind":"pith_short_12","alias_value":"IBK22QQUL5OZ","created_at":"2026-07-05T02:44:34.467457+00:00"},{"alias_kind":"pith_short_16","alias_value":"IBK22QQUL5OZ7RCA","created_at":"2026-07-05T02:44:34.467457+00:00"},{"alias_kind":"pith_short_8","alias_value":"IBK22QQU","created_at":"2026-07-05T02:44:34.467457+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.04276","citing_title":"Bridging Adaptivity and Safety: Learning Agile Collision-Free Locomotion Across Varied Physics","ref_index":39,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE","json":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE.json","graph_json":"https://pith.science/api/pith-number/IBK22QQUL5OZ7RCAKKHJP7DFOE/graph.json","events_json":"https://pith.science/api/pith-number/IBK22QQUL5OZ7RCAKKHJP7DFOE/events.json","paper":"https://pith.science/paper/IBK22QQU"},"agent_actions":{"view_html":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE","download_json":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE.json","view_paper":"https://pith.science/paper/IBK22QQU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2011.05869&json=true","fetch_graph":"https://pith.science/api/pith-number/IBK22QQUL5OZ7RCAKKHJP7DFOE/graph.json","fetch_events":"https://pith.science/api/pith-number/IBK22QQUL5OZ7RCAKKHJP7DFOE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE/action/storage_attestation","attest_author":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE/action/author_attestation","sign_citation":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE/action/citation_signature","submit_replication":"https://pith.science/pith/IBK22QQUL5OZ7RCAKKHJP7DFOE/action/replication_record"}},"created_at":"2026-07-05T02:44:34.467457+00:00","updated_at":"2026-07-05T02:44:34.467457+00:00"}