{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:GI7JMQ6K6UC4F7M3XTAILVZ6HV","short_pith_number":"pith:GI7JMQ6K","schema_version":"1.0","canonical_sha256":"323e9643caf505c2fd9bbcc085d73e3d4d6178a483955817d09811eeee12ea83","source":{"kind":"arxiv","id":"2212.10439","version":2},"attestation_state":"computed","paper":{"title":"Policy Gradient in Robust MDPs with Global Convergence Guarantee","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chin Pang Ho, Marek Petrik, Qiuhao Wang","submitted_at":"2022-12-20T17:14:14Z","abstract_excerpt":"Robust Markov decision processes (RMDPs) provide a promising framework for computing reliable policies in the face of model errors. Many successful reinforcement learning algorithms build on variations of policy-gradient methods, but adapting these methods to RMDPs has been challenging. As a result, the applicability of RMDPs to large, practical domains remains limited. This paper proposes a new Double-Loop Robust Policy Gradient (DRPG), the first generic policy gradient method for RMDPs. In contrast with prior robust policy gradient algorithms, DRPG monotonically reduces approximation errors "},"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":"2212.10439","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-12-20T17:14:14Z","cross_cats_sorted":[],"title_canon_sha256":"3c33a85efd4ddacaa17bb283eaa473c98efc2c4718b7dafc089ea0a71ad64cf2","abstract_canon_sha256":"340aa470b7b62da3ba854e6a0dbd1d334b86d5cab7abf3f2d8f78d1de480142a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:18:40.193033Z","signature_b64":"SgRaJv5g1jRMeQ+rY/cVxlyeBuP8qeA2f6Icrui094FnDwDIqStiKvWZ8PNJiXr7WiJ+EZtmqE6CAfkQN3bxBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"323e9643caf505c2fd9bbcc085d73e3d4d6178a483955817d09811eeee12ea83","last_reissued_at":"2026-07-05T08:18:40.192520Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:18:40.192520Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Policy Gradient in Robust MDPs with Global Convergence Guarantee","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chin Pang Ho, Marek Petrik, Qiuhao Wang","submitted_at":"2022-12-20T17:14:14Z","abstract_excerpt":"Robust Markov decision processes (RMDPs) provide a promising framework for computing reliable policies in the face of model errors. Many successful reinforcement learning algorithms build on variations of policy-gradient methods, but adapting these methods to RMDPs has been challenging. As a result, the applicability of RMDPs to large, practical domains remains limited. This paper proposes a new Double-Loop Robust Policy Gradient (DRPG), the first generic policy gradient method for RMDPs. In contrast with prior robust policy gradient algorithms, DRPG monotonically reduces approximation errors "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2212.10439","kind":"arxiv","version":2},"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/2212.10439/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":"2212.10439","created_at":"2026-07-05T08:18:40.192587+00:00"},{"alias_kind":"arxiv_version","alias_value":"2212.10439v2","created_at":"2026-07-05T08:18:40.192587+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2212.10439","created_at":"2026-07-05T08:18:40.192587+00:00"},{"alias_kind":"pith_short_12","alias_value":"GI7JMQ6K6UC4","created_at":"2026-07-05T08:18:40.192587+00:00"},{"alias_kind":"pith_short_16","alias_value":"GI7JMQ6K6UC4F7M3","created_at":"2026-07-05T08:18:40.192587+00:00"},{"alias_kind":"pith_short_8","alias_value":"GI7JMQ6K","created_at":"2026-07-05T08:18:40.192587+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2602.01505","citing_title":"Optimal Sample Complexity for Single Time-Scale Actor-Critic with Momentum","ref_index":55,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV","json":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV.json","graph_json":"https://pith.science/api/pith-number/GI7JMQ6K6UC4F7M3XTAILVZ6HV/graph.json","events_json":"https://pith.science/api/pith-number/GI7JMQ6K6UC4F7M3XTAILVZ6HV/events.json","paper":"https://pith.science/paper/GI7JMQ6K"},"agent_actions":{"view_html":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV","download_json":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV.json","view_paper":"https://pith.science/paper/GI7JMQ6K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2212.10439&json=true","fetch_graph":"https://pith.science/api/pith-number/GI7JMQ6K6UC4F7M3XTAILVZ6HV/graph.json","fetch_events":"https://pith.science/api/pith-number/GI7JMQ6K6UC4F7M3XTAILVZ6HV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV/action/storage_attestation","attest_author":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV/action/author_attestation","sign_citation":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV/action/citation_signature","submit_replication":"https://pith.science/pith/GI7JMQ6K6UC4F7M3XTAILVZ6HV/action/replication_record"}},"created_at":"2026-07-05T08:18:40.192587+00:00","updated_at":"2026-07-05T08:18:40.192587+00:00"}