{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:DTUPU4YPWBVOSXHAUETU2SFEGN","short_pith_number":"pith:DTUPU4YP","schema_version":"1.0","canonical_sha256":"1ce8fa730fb06ae95ce0a1274d48a43342064e3026b1f2dac2fc69a8a17b28f2","source":{"kind":"arxiv","id":"2605.10236","version":3},"attestation_state":"computed","paper":{"title":"When Does Non-Uniform Replay Matter in Reinforcement Learning?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Non-uniform replay improves reinforcement learning sample efficiency mainly when replay volume is low, provided sampling entropy stays high.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Michal Korniak, Michal Nauman, Miko{\\l}aj Czarnecki, Pieter Abbeel, Piotr Mi{\\l}o\\'s, Yarden As","submitted_at":"2026-05-11T09:11:05Z","abstract_excerpt":"Modern off-policy reinforcement learning algorithms often rely on simple uniform replay sampling and it remains unclear when and why non-uniform replay improves over this strong baseline. Across diverse RL settings, we show that the effectiveness of non-uniform replay is governed by three factors: replay volume, the number of replayed transitions per environment step; expected recency, how recent sampled transitions are; and the entropy of the replay sampling distribution. Our main contribution is clarifying when non-uniform replay is beneficial and providing practical guidance for replay desi"},"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":"2605.10236","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-11T09:11:05Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c552ff302da0820dc38cc0a64a17100907ea6d70298f50f5faaf1c656cedb6a3","abstract_canon_sha256":"620bf1ffe50dbaacaa7a9d4a62ea8c89e696204990ecc4b3b8aa6e570113787f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:04:35.666398Z","signature_b64":"qzaHc4gXZVKgOLcwqMGaKl1ZEUqF6VdkviiQmR6/S1dOOwJxs/f2TEQRZMVeHWLkBhTX6S/qJaHzZQuBzEk5Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1ce8fa730fb06ae95ce0a1274d48a43342064e3026b1f2dac2fc69a8a17b28f2","last_reissued_at":"2026-05-20T00:04:35.665186Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:04:35.665186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When Does Non-Uniform Replay Matter in Reinforcement Learning?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Non-uniform replay improves reinforcement learning sample efficiency mainly when replay volume is low, provided sampling entropy stays high.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Michal Korniak, Michal Nauman, Miko{\\l}aj Czarnecki, Pieter Abbeel, Piotr Mi{\\l}o\\'s, Yarden As","submitted_at":"2026-05-11T09:11:05Z","abstract_excerpt":"Modern off-policy reinforcement learning algorithms often rely on simple uniform replay sampling and it remains unclear when and why non-uniform replay improves over this strong baseline. Across diverse RL settings, we show that the effectiveness of non-uniform replay is governed by three factors: replay volume, the number of replayed transitions per environment step; expected recency, how recent sampled transitions are; and the entropy of the replay sampling distribution. Our main contribution is clarifying when non-uniform replay is beneficial and providing practical guidance for replay desi"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"the effectiveness of non-uniform replay is governed by three factors: replay volume, the number of replayed transitions per environment step; expected recency, how recent sampled transitions are; and the entropy of the replay sampling distribution. ... non-uniform replay is most beneficial when replay volume is low, and that high-entropy sampling is important even at comparable expected recency.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the three identified factors comprehensively govern non-uniform replay effectiveness and that the observed benefits will generalize beyond the specific algorithms, benchmarks, and parallel-simulation regimes tested.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Non-uniform replay improves RL sample efficiency mainly in low replay-volume regimes, with high-entropy sampling being key even at comparable recency.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Non-uniform replay improves reinforcement learning sample efficiency mainly when replay volume is low, provided sampling entropy stays high.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"43df73cedae58c20bca3de2933d49bbd268b878f2e44eb1c7fa06e009cb362a9"},"source":{"id":"2605.10236","kind":"arxiv","version":3},"verdict":{"id":"7d7639d2-4b3a-4ee5-a83f-d74534203b2e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T06:22:47.019414Z","strongest_claim":"the effectiveness of non-uniform replay is governed by three factors: replay volume, the number of replayed transitions per environment step; expected recency, how recent sampled transitions are; and the entropy of the replay sampling distribution. ... non-uniform replay is most beneficial when replay volume is low, and that high-entropy sampling is important even at comparable expected recency.","one_line_summary":"Non-uniform replay improves RL sample efficiency mainly in low replay-volume regimes, with high-entropy sampling being key even at comparable recency.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the three identified factors comprehensively govern non-uniform replay effectiveness and that the observed benefits will generalize beyond the specific algorithms, benchmarks, and parallel-simulation regimes tested.","pith_extraction_headline":"Non-uniform replay improves reinforcement learning sample efficiency mainly when replay volume is low, provided sampling entropy stays high."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.10236/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-19T15:38:10.202722Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T11:31:19.746771Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T09:33:47.881786Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"9ad6adbe4ce54f729447c63461f7ce37017431aba2fc990d1fa3e8848f8c5734"},"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":"2605.10236","created_at":"2026-05-20T00:04:35.665385+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.10236v3","created_at":"2026-05-20T00:04:35.665385+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.10236","created_at":"2026-05-20T00:04:35.665385+00:00"},{"alias_kind":"pith_short_12","alias_value":"DTUPU4YPWBVO","created_at":"2026-05-20T00:04:35.665385+00:00"},{"alias_kind":"pith_short_16","alias_value":"DTUPU4YPWBVOSXHA","created_at":"2026-05-20T00:04:35.665385+00:00"},{"alias_kind":"pith_short_8","alias_value":"DTUPU4YP","created_at":"2026-05-20T00:04:35.665385+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/DTUPU4YPWBVOSXHAUETU2SFEGN","json":"https://pith.science/pith/DTUPU4YPWBVOSXHAUETU2SFEGN.json","graph_json":"https://pith.science/api/pith-number/DTUPU4YPWBVOSXHAUETU2SFEGN/graph.json","events_json":"https://pith.science/api/pith-number/DTUPU4YPWBVOSXHAUETU2SFEGN/events.json","paper":"https://pith.science/paper/DTUPU4YP"},"agent_actions":{"view_html":"https://pith.science/pith/DTUPU4YPWBVOSXHAUETU2SFEGN","download_json":"https://pith.science/pith/DTUPU4YPWBVOSXHAUETU2SFEGN.json","view_paper":"https://pith.science/paper/DTUPU4YP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.10236&json=true","fetch_graph":"https://pith.science/api/pith-number/DTUPU4YPWBVOSXHAUETU2SFEGN/graph.json","fetch_events":"https://pith.science/api/pith-number/DTUPU4YPWBVOSXHAUETU2SFEGN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DTUPU4YPWBVOSXHAUETU2SFEGN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DTUPU4YPWBVOSXHAUETU2SFEGN/action/storage_attestation","attest_author":"https://pith.science/pith/DTUPU4YPWBVOSXHAUETU2SFEGN/action/author_attestation","sign_citation":"https://pith.science/pith/DTUPU4YPWBVOSXHAUETU2SFEGN/action/citation_signature","submit_replication":"https://pith.science/pith/DTUPU4YPWBVOSXHAUETU2SFEGN/action/replication_record"}},"created_at":"2026-05-20T00:04:35.665385+00:00","updated_at":"2026-05-20T00:04:35.665385+00:00"}