{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:JUZXUANQK2CHCPBQM3WEXK6WXW","short_pith_number":"pith:JUZXUANQ","schema_version":"1.0","canonical_sha256":"4d337a01b05684713c3066ec4babd6bd940aa51100e6a8e8f6bc0139d684e471","source":{"kind":"arxiv","id":"1706.00387","version":1},"attestation_state":"computed","paper":{"title":"Interpolated Policy Gradient: Merging On-Policy and Off-Policy Gradient Estimation for Deep Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.RO"],"primary_cat":"cs.LG","authors_text":"Bernhard Sch\\\"olkopf, Richard E. Turner, Sergey Levine, Shixiang Gu, Timothy Lillicrap, Zoubin Ghahramani","submitted_at":"2017-06-01T17:00:52Z","abstract_excerpt":"Off-policy model-free deep reinforcement learning methods using previously collected data can improve sample efficiency over on-policy policy gradient techniques. On the other hand, on-policy algorithms are often more stable and easier to use. This paper examines, both theoretically and empirically, approaches to merging on- and off-policy updates for deep reinforcement learning. Theoretical results show that off-policy updates with a value function estimator can be interpolated with on-policy policy gradient updates whilst still satisfying performance bounds. Our analysis uses control variate"},"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":"1706.00387","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-06-01T17:00:52Z","cross_cats_sorted":["cs.AI","cs.RO"],"title_canon_sha256":"d5099d2f721e787759304db2d8511a3078ab0a1f5a8c8fa1cbe07826afa21fdc","abstract_canon_sha256":"441da829df57dc6938085fd89154a1fb895c875e97bc1846e86ca17c60bfb356"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:43:14.162997Z","signature_b64":"0ShOXg1HoU9dBjaoxHpOCRldlDIawwnU7B+9fIKk7L5oMkIhCcUBV6XXTYmiAaSlfDynEI1ycta0xrXxjQwNDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4d337a01b05684713c3066ec4babd6bd940aa51100e6a8e8f6bc0139d684e471","last_reissued_at":"2026-05-18T00:43:14.162296Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:43:14.162296Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Interpolated Policy Gradient: Merging On-Policy and Off-Policy Gradient Estimation for Deep Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.RO"],"primary_cat":"cs.LG","authors_text":"Bernhard Sch\\\"olkopf, Richard E. Turner, Sergey Levine, Shixiang Gu, Timothy Lillicrap, Zoubin Ghahramani","submitted_at":"2017-06-01T17:00:52Z","abstract_excerpt":"Off-policy model-free deep reinforcement learning methods using previously collected data can improve sample efficiency over on-policy policy gradient techniques. On the other hand, on-policy algorithms are often more stable and easier to use. This paper examines, both theoretically and empirically, approaches to merging on- and off-policy updates for deep reinforcement learning. Theoretical results show that off-policy updates with a value function estimator can be interpolated with on-policy policy gradient updates whilst still satisfying performance bounds. Our analysis uses control variate"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1706.00387","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":""},"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":"1706.00387","created_at":"2026-05-18T00:43:14.162431+00:00"},{"alias_kind":"arxiv_version","alias_value":"1706.00387v1","created_at":"2026-05-18T00:43:14.162431+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1706.00387","created_at":"2026-05-18T00:43:14.162431+00:00"},{"alias_kind":"pith_short_12","alias_value":"JUZXUANQK2CH","created_at":"2026-05-18T12:31:24.725408+00:00"},{"alias_kind":"pith_short_16","alias_value":"JUZXUANQK2CHCPBQ","created_at":"2026-05-18T12:31:24.725408+00:00"},{"alias_kind":"pith_short_8","alias_value":"JUZXUANQ","created_at":"2026-05-18T12:31:24.725408+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2603.12031","citing_title":"AGMARL-DKS: An Adaptive Graph-Enhanced Multi-Agent Reinforcement Learning for Dynamic Kubernetes Scheduling","ref_index":63,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW","json":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW.json","graph_json":"https://pith.science/api/pith-number/JUZXUANQK2CHCPBQM3WEXK6WXW/graph.json","events_json":"https://pith.science/api/pith-number/JUZXUANQK2CHCPBQM3WEXK6WXW/events.json","paper":"https://pith.science/paper/JUZXUANQ"},"agent_actions":{"view_html":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW","download_json":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW.json","view_paper":"https://pith.science/paper/JUZXUANQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1706.00387&json=true","fetch_graph":"https://pith.science/api/pith-number/JUZXUANQK2CHCPBQM3WEXK6WXW/graph.json","fetch_events":"https://pith.science/api/pith-number/JUZXUANQK2CHCPBQM3WEXK6WXW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW/action/storage_attestation","attest_author":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW/action/author_attestation","sign_citation":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW/action/citation_signature","submit_replication":"https://pith.science/pith/JUZXUANQK2CHCPBQM3WEXK6WXW/action/replication_record"}},"created_at":"2026-05-18T00:43:14.162431+00:00","updated_at":"2026-05-18T00:43:14.162431+00:00"}