{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:NDKWDS7EIAT72WT2CDXCGBQDUB","short_pith_number":"pith:NDKWDS7E","schema_version":"1.0","canonical_sha256":"68d561cbe44027fd5a7a10ee230603a07a9bedab4faa5ef320fab93b373d7160","source":{"kind":"arxiv","id":"2104.13446","version":2},"attestation_state":"computed","paper":{"title":"Semi-On-Policy Training for Sample Efficient Multi-Agent Policy Gradients","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.MA"],"primary_cat":"cs.LG","authors_text":"Bei Peng, Bozhidar Vasilev, Shimon Whiteson, Tarun Gupta","submitted_at":"2021-04-27T19:37:01Z","abstract_excerpt":"Policy gradient methods are an attractive approach to multi-agent reinforcement learning problems due to their convergence properties and robustness in partially observable scenarios. However, there is a significant performance gap between state-of-the-art policy gradient and value-based methods on the popular StarCraft Multi-Agent Challenge (SMAC) benchmark. In this paper, we introduce semi-on-policy (SOP) training as an effective and computationally efficient way to address the sample inefficiency of on-policy policy gradient methods. We enhance two state-of-the-art policy gradient algorithm"},"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":"2104.13446","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-04-27T19:37:01Z","cross_cats_sorted":["cs.MA"],"title_canon_sha256":"20281de520c13844c317b32224b06a6910c72aab76be10c4cee1a7a2b22d249f","abstract_canon_sha256":"7928971303258b289c3c2b9c9a2d201ccfeca94748487256552e8f8f8736ca1e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:38:09.621023Z","signature_b64":"TaZeVFVe9jSeM/gK/iyn04MNAonIF2F1gOSnPgmaPpr8/LF6S9KUbb+YcnbjFY2FUIF75T9JlX2rAZzZHqjLDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"68d561cbe44027fd5a7a10ee230603a07a9bedab4faa5ef320fab93b373d7160","last_reissued_at":"2026-07-05T02:38:09.620554Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:38:09.620554Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Semi-On-Policy Training for Sample Efficient Multi-Agent Policy Gradients","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.MA"],"primary_cat":"cs.LG","authors_text":"Bei Peng, Bozhidar Vasilev, Shimon Whiteson, Tarun Gupta","submitted_at":"2021-04-27T19:37:01Z","abstract_excerpt":"Policy gradient methods are an attractive approach to multi-agent reinforcement learning problems due to their convergence properties and robustness in partially observable scenarios. However, there is a significant performance gap between state-of-the-art policy gradient and value-based methods on the popular StarCraft Multi-Agent Challenge (SMAC) benchmark. In this paper, we introduce semi-on-policy (SOP) training as an effective and computationally efficient way to address the sample inefficiency of on-policy policy gradient methods. We enhance two state-of-the-art policy gradient algorithm"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.13446","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/2104.13446/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":"2104.13446","created_at":"2026-07-05T02:38:09.620613+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.13446v2","created_at":"2026-07-05T02:38:09.620613+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.13446","created_at":"2026-07-05T02:38:09.620613+00:00"},{"alias_kind":"pith_short_12","alias_value":"NDKWDS7EIAT7","created_at":"2026-07-05T02:38:09.620613+00:00"},{"alias_kind":"pith_short_16","alias_value":"NDKWDS7EIAT72WT2","created_at":"2026-07-05T02:38:09.620613+00:00"},{"alias_kind":"pith_short_8","alias_value":"NDKWDS7E","created_at":"2026-07-05T02:38:09.620613+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB","json":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB.json","graph_json":"https://pith.science/api/pith-number/NDKWDS7EIAT72WT2CDXCGBQDUB/graph.json","events_json":"https://pith.science/api/pith-number/NDKWDS7EIAT72WT2CDXCGBQDUB/events.json","paper":"https://pith.science/paper/NDKWDS7E"},"agent_actions":{"view_html":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB","download_json":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB.json","view_paper":"https://pith.science/paper/NDKWDS7E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.13446&json=true","fetch_graph":"https://pith.science/api/pith-number/NDKWDS7EIAT72WT2CDXCGBQDUB/graph.json","fetch_events":"https://pith.science/api/pith-number/NDKWDS7EIAT72WT2CDXCGBQDUB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB/action/storage_attestation","attest_author":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB/action/author_attestation","sign_citation":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB/action/citation_signature","submit_replication":"https://pith.science/pith/NDKWDS7EIAT72WT2CDXCGBQDUB/action/replication_record"}},"created_at":"2026-07-05T02:38:09.620613+00:00","updated_at":"2026-07-05T02:38:09.620613+00:00"}