{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VW3SKHYIRILFLDMQSFFQYQWKSV","short_pith_number":"pith:VW3SKHYI","schema_version":"1.0","canonical_sha256":"adb7251f088a16558d90914b0c42ca957589688be7daa466cc912184522f40dc","source":{"kind":"arxiv","id":"2501.15034","version":1},"attestation_state":"computed","paper":{"title":"Divergence-Augmented Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jiechao Xiong, Qing Wang, Tong Zhang, Yingru Li","submitted_at":"2025-01-25T02:35:46Z","abstract_excerpt":"In deep reinforcement learning, policy optimization methods need to deal with issues such as function approximation and the reuse of off-policy data. Standard policy gradient methods do not handle off-policy data well, leading to premature convergence and instability. This paper introduces a method to stabilize policy optimization when off-policy data are reused. The idea is to include a Bregman divergence between the behavior policy that generates the data and the current policy to ensure small and safe policy updates with off-policy data. The Bregman divergence is calculated between the stat"},"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":"2501.15034","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-25T02:35:46Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"940d581a2b5b68f81ab003fd266b53dbe8b4a5dd67507251b1e0fd880ce0bb12","abstract_canon_sha256":"be5fb3a85818a5fc55f7c8beaa5063d66afb7f1f46f2875a3a3d56be657fd2a7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:05:22.451478Z","signature_b64":"Nzo7lZUYOmGluVUziUWDAu7/ykAuEyEg6lf/s1LsJ/voEjo5zC3noKx0WvtoQkunOoF9QOIvVTgWx8/mhjzwBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"adb7251f088a16558d90914b0c42ca957589688be7daa466cc912184522f40dc","last_reissued_at":"2026-07-05T10:05:22.450950Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:05:22.450950Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Divergence-Augmented Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jiechao Xiong, Qing Wang, Tong Zhang, Yingru Li","submitted_at":"2025-01-25T02:35:46Z","abstract_excerpt":"In deep reinforcement learning, policy optimization methods need to deal with issues such as function approximation and the reuse of off-policy data. Standard policy gradient methods do not handle off-policy data well, leading to premature convergence and instability. This paper introduces a method to stabilize policy optimization when off-policy data are reused. The idea is to include a Bregman divergence between the behavior policy that generates the data and the current policy to ensure small and safe policy updates with off-policy data. The Bregman divergence is calculated between the stat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.15034","kind":"arxiv","version":1},"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/2501.15034/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":"2501.15034","created_at":"2026-07-05T10:05:22.451015+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.15034v1","created_at":"2026-07-05T10:05:22.451015+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.15034","created_at":"2026-07-05T10:05:22.451015+00:00"},{"alias_kind":"pith_short_12","alias_value":"VW3SKHYIRILF","created_at":"2026-07-05T10:05:22.451015+00:00"},{"alias_kind":"pith_short_16","alias_value":"VW3SKHYIRILFLDMQ","created_at":"2026-07-05T10:05:22.451015+00:00"},{"alias_kind":"pith_short_8","alias_value":"VW3SKHYI","created_at":"2026-07-05T10:05:22.451015+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.15034","citing_title":"Divergence-Augmented Policy Optimization","ref_index":18,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV","json":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV.json","graph_json":"https://pith.science/api/pith-number/VW3SKHYIRILFLDMQSFFQYQWKSV/graph.json","events_json":"https://pith.science/api/pith-number/VW3SKHYIRILFLDMQSFFQYQWKSV/events.json","paper":"https://pith.science/paper/VW3SKHYI"},"agent_actions":{"view_html":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV","download_json":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV.json","view_paper":"https://pith.science/paper/VW3SKHYI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.15034&json=true","fetch_graph":"https://pith.science/api/pith-number/VW3SKHYIRILFLDMQSFFQYQWKSV/graph.json","fetch_events":"https://pith.science/api/pith-number/VW3SKHYIRILFLDMQSFFQYQWKSV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV/action/storage_attestation","attest_author":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV/action/author_attestation","sign_citation":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV/action/citation_signature","submit_replication":"https://pith.science/pith/VW3SKHYIRILFLDMQSFFQYQWKSV/action/replication_record"}},"created_at":"2026-07-05T10:05:22.451015+00:00","updated_at":"2026-07-05T10:05:22.451015+00:00"}