{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:PNAEISQ4LRE4Z6WBODGW7OTHU4","short_pith_number":"pith:PNAEISQ4","schema_version":"1.0","canonical_sha256":"7b40444a1c5c49ccfac170cd6fba67a70dfdda90d4a1f5e082c0ea68f8cd1668","source":{"kind":"arxiv","id":"2003.04302","version":1},"attestation_state":"computed","paper":{"title":"Stochastic Recursive Momentum for Policy Gradient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Huizhuo Yuan, Ji Liu, Xiangru Lian, Yuren Zhou","submitted_at":"2020-03-09T17:59:03Z","abstract_excerpt":"In this paper, we propose a novel algorithm named STOchastic Recursive Momentum for Policy Gradient (STORM-PG), which operates a SARAH-type stochastic recursive variance-reduced policy gradient in an exponential moving average fashion. STORM-PG enjoys a provably sharp $O(1/\\epsilon^3)$ sample complexity bound for STORM-PG, matching the best-known convergence rate for policy gradient algorithm. In the mean time, STORM-PG avoids the alternations between large batches and small batches which persists in comparable variance-reduced policy gradient methods, allowing considerably simpler parameter t"},"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":"2003.04302","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2020-03-09T17:59:03Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"b52dd2d910a8279b93cf3b39cd00689e4c994002bf700c6b14af6c402391953f","abstract_canon_sha256":"d3e57597f92ae92eac85f7140fa2b49160817614c111339f6836c896e5599be2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:46:41.308583Z","signature_b64":"NEr1M/Sfvwbyh18zyuzHOQLM9iz115qfUSr4zCNOW1BRwLyumPxeGnH/CU3bZzZElij5D/nBeQATShFVgiFdBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7b40444a1c5c49ccfac170cd6fba67a70dfdda90d4a1f5e082c0ea68f8cd1668","last_reissued_at":"2026-07-05T00:46:41.308213Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:46:41.308213Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Stochastic Recursive Momentum for Policy Gradient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Huizhuo Yuan, Ji Liu, Xiangru Lian, Yuren Zhou","submitted_at":"2020-03-09T17:59:03Z","abstract_excerpt":"In this paper, we propose a novel algorithm named STOchastic Recursive Momentum for Policy Gradient (STORM-PG), which operates a SARAH-type stochastic recursive variance-reduced policy gradient in an exponential moving average fashion. STORM-PG enjoys a provably sharp $O(1/\\epsilon^3)$ sample complexity bound for STORM-PG, matching the best-known convergence rate for policy gradient algorithm. In the mean time, STORM-PG avoids the alternations between large batches and small batches which persists in comparable variance-reduced policy gradient methods, allowing considerably simpler parameter t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2003.04302","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/2003.04302/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":"2003.04302","created_at":"2026-07-05T00:46:41.308270+00:00"},{"alias_kind":"arxiv_version","alias_value":"2003.04302v1","created_at":"2026-07-05T00:46:41.308270+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2003.04302","created_at":"2026-07-05T00:46:41.308270+00:00"},{"alias_kind":"pith_short_12","alias_value":"PNAEISQ4LRE4","created_at":"2026-07-05T00:46:41.308270+00:00"},{"alias_kind":"pith_short_16","alias_value":"PNAEISQ4LRE4Z6WB","created_at":"2026-07-05T00:46:41.308270+00:00"},{"alias_kind":"pith_short_8","alias_value":"PNAEISQ4","created_at":"2026-07-05T00:46:41.308270+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.06178","citing_title":"Reusing Trajectories in Policy Gradients Enables Fast Convergence","ref_index":551,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4","json":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4.json","graph_json":"https://pith.science/api/pith-number/PNAEISQ4LRE4Z6WBODGW7OTHU4/graph.json","events_json":"https://pith.science/api/pith-number/PNAEISQ4LRE4Z6WBODGW7OTHU4/events.json","paper":"https://pith.science/paper/PNAEISQ4"},"agent_actions":{"view_html":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4","download_json":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4.json","view_paper":"https://pith.science/paper/PNAEISQ4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2003.04302&json=true","fetch_graph":"https://pith.science/api/pith-number/PNAEISQ4LRE4Z6WBODGW7OTHU4/graph.json","fetch_events":"https://pith.science/api/pith-number/PNAEISQ4LRE4Z6WBODGW7OTHU4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4/action/storage_attestation","attest_author":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4/action/author_attestation","sign_citation":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4/action/citation_signature","submit_replication":"https://pith.science/pith/PNAEISQ4LRE4Z6WBODGW7OTHU4/action/replication_record"}},"created_at":"2026-07-05T00:46:41.308270+00:00","updated_at":"2026-07-05T00:46:41.308270+00:00"}