{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HTYLDZNQRRZELXNQRGXV4A6XVE","short_pith_number":"pith:HTYLDZNQ","schema_version":"1.0","canonical_sha256":"3cf0b1e5b08c7245ddb089af5e03d7a93a2513799afc4441eb2d769805c00ad4","source":{"kind":"arxiv","id":"2405.16383","version":1},"attestation_state":"computed","paper":{"title":"Rewarded Region Replay (R3) for Policy Learning with Discrete Action Space","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bangzheng Li, Ningshan Ma, Zifan Wang","submitted_at":"2024-05-26T00:01:29Z","abstract_excerpt":"We introduce a new on-policy algorithm called Rewarded Region Replay (R3), which significantly improves on PPO in solving environments with discrete action spaces. R3 improves sample efficiency by using a replay buffer which contains past successful trajectories with reward above a certain threshold, which are used to update a PPO agent with importance sampling. Crucially, we discard the importance sampling factors which are above a certain ratio to reduce variance and stabilize training. We found that R3 significantly outperforms PPO in Minigrid environments with sparse rewards and discrete a"},"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":"2405.16383","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-26T00:01:29Z","cross_cats_sorted":[],"title_canon_sha256":"57014655e767efa13a1734b880571b9f111dfc7bfdb91918eddb176eb241711d","abstract_canon_sha256":"196a71a5eb4fe82975cbddf182cb32a0a6b272662ef0a22184b5dacbc9cc9751"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:23:29.400365Z","signature_b64":"Vaei7NnAvJw/dmFuDVJYXUCFb+ejhNjjPiy4Mjw2NJGXBcD8ccfdZrloXidTPzfpihDlzmDYvM+C/+u2QBY4Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3cf0b1e5b08c7245ddb089af5e03d7a93a2513799afc4441eb2d769805c00ad4","last_reissued_at":"2026-07-05T08:23:29.399904Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:23:29.399904Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rewarded Region Replay (R3) for Policy Learning with Discrete Action Space","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bangzheng Li, Ningshan Ma, Zifan Wang","submitted_at":"2024-05-26T00:01:29Z","abstract_excerpt":"We introduce a new on-policy algorithm called Rewarded Region Replay (R3), which significantly improves on PPO in solving environments with discrete action spaces. R3 improves sample efficiency by using a replay buffer which contains past successful trajectories with reward above a certain threshold, which are used to update a PPO agent with importance sampling. Crucially, we discard the importance sampling factors which are above a certain ratio to reduce variance and stabilize training. We found that R3 significantly outperforms PPO in Minigrid environments with sparse rewards and discrete a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.16383","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/2405.16383/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":"2405.16383","created_at":"2026-07-05T08:23:29.399961+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.16383v1","created_at":"2026-07-05T08:23:29.399961+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.16383","created_at":"2026-07-05T08:23:29.399961+00:00"},{"alias_kind":"pith_short_12","alias_value":"HTYLDZNQRRZE","created_at":"2026-07-05T08:23:29.399961+00:00"},{"alias_kind":"pith_short_16","alias_value":"HTYLDZNQRRZELXNQ","created_at":"2026-07-05T08:23:29.399961+00:00"},{"alias_kind":"pith_short_8","alias_value":"HTYLDZNQ","created_at":"2026-07-05T08:23:29.399961+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/HTYLDZNQRRZELXNQRGXV4A6XVE","json":"https://pith.science/pith/HTYLDZNQRRZELXNQRGXV4A6XVE.json","graph_json":"https://pith.science/api/pith-number/HTYLDZNQRRZELXNQRGXV4A6XVE/graph.json","events_json":"https://pith.science/api/pith-number/HTYLDZNQRRZELXNQRGXV4A6XVE/events.json","paper":"https://pith.science/paper/HTYLDZNQ"},"agent_actions":{"view_html":"https://pith.science/pith/HTYLDZNQRRZELXNQRGXV4A6XVE","download_json":"https://pith.science/pith/HTYLDZNQRRZELXNQRGXV4A6XVE.json","view_paper":"https://pith.science/paper/HTYLDZNQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.16383&json=true","fetch_graph":"https://pith.science/api/pith-number/HTYLDZNQRRZELXNQRGXV4A6XVE/graph.json","fetch_events":"https://pith.science/api/pith-number/HTYLDZNQRRZELXNQRGXV4A6XVE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HTYLDZNQRRZELXNQRGXV4A6XVE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HTYLDZNQRRZELXNQRGXV4A6XVE/action/storage_attestation","attest_author":"https://pith.science/pith/HTYLDZNQRRZELXNQRGXV4A6XVE/action/author_attestation","sign_citation":"https://pith.science/pith/HTYLDZNQRRZELXNQRGXV4A6XVE/action/citation_signature","submit_replication":"https://pith.science/pith/HTYLDZNQRRZELXNQRGXV4A6XVE/action/replication_record"}},"created_at":"2026-07-05T08:23:29.399961+00:00","updated_at":"2026-07-05T08:23:29.399961+00:00"}