{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:VLYUJEJZIZ3UAVXBUSK2U5MLZQ","short_pith_number":"pith:VLYUJEJZ","schema_version":"1.0","canonical_sha256":"aaf144913946774056e1a495aa758bcc3ab7228054b6e1e79d30ea952960f336","source":{"kind":"arxiv","id":"2210.04157","version":1},"attestation_state":"computed","paper":{"title":"The Role of Coverage in Online Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Dylan J. Foster, Nan Jiang, Sham M. Kakade, Tengyang Xie, Yu Bai","submitted_at":"2022-10-09T03:50:05Z","abstract_excerpt":"Coverage conditions -- which assert that the data logging distribution adequately covers the state space -- play a fundamental role in determining the sample complexity of offline reinforcement learning. While such conditions might seem irrelevant to online reinforcement learning at first glance, we establish a new connection by showing -- somewhat surprisingly -- that the mere existence of a data distribution with good coverage can enable sample-efficient online RL. Concretely, we show that coverability -- that is, existence of a data distribution that satisfies a ubiquitous coverage conditio"},"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":"2210.04157","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-10-09T03:50:05Z","cross_cats_sorted":["cs.AI","math.OC","stat.ML"],"title_canon_sha256":"aa4f44cd6a444e4d81c23eee7751df9d263ce9688554c9007471846f303fc45b","abstract_canon_sha256":"f08c3bb04d9283538ed9b5048b00498e75c8c9b1425eb44bd21bdce621bbe3a2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:04:36.895996Z","signature_b64":"KXgbfaHCZrqocv2CEXusQJ9pj41OPyF5hCpkZd9xKEVYxmdSXBogItuxWNwHF7rLeP+ooZ0Ytujs7gnSUsi8CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aaf144913946774056e1a495aa758bcc3ab7228054b6e1e79d30ea952960f336","last_reissued_at":"2026-07-05T05:04:36.895506Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:04:36.895506Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Role of Coverage in Online Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Dylan J. Foster, Nan Jiang, Sham M. Kakade, Tengyang Xie, Yu Bai","submitted_at":"2022-10-09T03:50:05Z","abstract_excerpt":"Coverage conditions -- which assert that the data logging distribution adequately covers the state space -- play a fundamental role in determining the sample complexity of offline reinforcement learning. While such conditions might seem irrelevant to online reinforcement learning at first glance, we establish a new connection by showing -- somewhat surprisingly -- that the mere existence of a data distribution with good coverage can enable sample-efficient online RL. Concretely, we show that coverability -- that is, existence of a data distribution that satisfies a ubiquitous coverage conditio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.04157","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/2210.04157/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":"2210.04157","created_at":"2026-07-05T05:04:36.895559+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.04157v1","created_at":"2026-07-05T05:04:36.895559+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.04157","created_at":"2026-07-05T05:04:36.895559+00:00"},{"alias_kind":"pith_short_12","alias_value":"VLYUJEJZIZ3U","created_at":"2026-07-05T05:04:36.895559+00:00"},{"alias_kind":"pith_short_16","alias_value":"VLYUJEJZIZ3UAVXB","created_at":"2026-07-05T05:04:36.895559+00:00"},{"alias_kind":"pith_short_8","alias_value":"VLYUJEJZ","created_at":"2026-07-05T05:04:36.895559+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.05375","citing_title":"Fitted Occupancy-Ratio Evaluation without Bellman Completeness","ref_index":90,"is_internal_anchor":true},{"citing_arxiv_id":"2606.20107","citing_title":"Quantile of Means: A Bonus-Free Ensemble Method for Minimax Optimal Reinforcement Learning","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06053","citing_title":"Online KL-Regularized Reinforcement Learning with Function Approximation under Misspecification","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07049","citing_title":"Towards Differentially Private Reinforcement Learning with General Function Approximation","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ","json":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ.json","graph_json":"https://pith.science/api/pith-number/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/graph.json","events_json":"https://pith.science/api/pith-number/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/events.json","paper":"https://pith.science/paper/VLYUJEJZ"},"agent_actions":{"view_html":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ","download_json":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ.json","view_paper":"https://pith.science/paper/VLYUJEJZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.04157&json=true","fetch_graph":"https://pith.science/api/pith-number/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/graph.json","fetch_events":"https://pith.science/api/pith-number/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/action/storage_attestation","attest_author":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/action/author_attestation","sign_citation":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/action/citation_signature","submit_replication":"https://pith.science/pith/VLYUJEJZIZ3UAVXBUSK2U5MLZQ/action/replication_record"}},"created_at":"2026-07-05T05:04:36.895559+00:00","updated_at":"2026-07-05T05:04:36.895559+00:00"}