{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:PHDLNRYIW7QMJ67CQGWD74P3XZ","short_pith_number":"pith:PHDLNRYI","schema_version":"1.0","canonical_sha256":"79c6b6c708b7e0c4fbe281ac3ff1fbbe6a8f22a68e22ff6f917ddd3c17e750e5","source":{"kind":"arxiv","id":"2211.11092","version":2},"attestation_state":"computed","paper":{"title":"Q-Ensemble for Offline RL: Don't Scale the Ensemble, Scale the Batch Size","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.NE"],"primary_cat":"cs.LG","authors_text":"Alexander Nikulin, Denis Tarasov, Dmitry Akimov, Sergey Kolesnikov, Vladislav Kurenkov","submitted_at":"2022-11-20T21:48:25Z","abstract_excerpt":"Training large neural networks is known to be time-consuming, with the learning duration taking days or even weeks. To address this problem, large-batch optimization was introduced. This approach demonstrated that scaling mini-batch sizes with appropriate learning rate adjustments can speed up the training process by orders of magnitude. While long training time was not typically a major issue for model-free deep offline RL algorithms, recently introduced Q-ensemble methods achieving state-of-the-art performance made this issue more relevant, notably extending the training duration. In this wo"},"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":"2211.11092","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-11-20T21:48:25Z","cross_cats_sorted":["cs.AI","cs.NE"],"title_canon_sha256":"d74a4b78303942bbda261db9826f04ebc1c8fcac610aad3519fc5b2135e3640a","abstract_canon_sha256":"8105126d22dc92eab5119a5e4a9e0070c80f786ef2a03c73a320fdab79412b8d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:36:51.156220Z","signature_b64":"HNE5h+avqIG59WjmOc8znjSuK+DImfbfFnwDRqh0eGswvGCqaK6FfwbKEY+jnGSUURCi0u1slNmLgQjX2n2FBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"79c6b6c708b7e0c4fbe281ac3ff1fbbe6a8f22a68e22ff6f917ddd3c17e750e5","last_reissued_at":"2026-07-05T05:36:51.155753Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:36:51.155753Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Q-Ensemble for Offline RL: Don't Scale the Ensemble, Scale the Batch Size","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.NE"],"primary_cat":"cs.LG","authors_text":"Alexander Nikulin, Denis Tarasov, Dmitry Akimov, Sergey Kolesnikov, Vladislav Kurenkov","submitted_at":"2022-11-20T21:48:25Z","abstract_excerpt":"Training large neural networks is known to be time-consuming, with the learning duration taking days or even weeks. To address this problem, large-batch optimization was introduced. This approach demonstrated that scaling mini-batch sizes with appropriate learning rate adjustments can speed up the training process by orders of magnitude. While long training time was not typically a major issue for model-free deep offline RL algorithms, recently introduced Q-ensemble methods achieving state-of-the-art performance made this issue more relevant, notably extending the training duration. In this wo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.11092","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/2211.11092/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":"2211.11092","created_at":"2026-07-05T05:36:51.155815+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.11092v2","created_at":"2026-07-05T05:36:51.155815+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.11092","created_at":"2026-07-05T05:36:51.155815+00:00"},{"alias_kind":"pith_short_12","alias_value":"PHDLNRYIW7QM","created_at":"2026-07-05T05:36:51.155815+00:00"},{"alias_kind":"pith_short_16","alias_value":"PHDLNRYIW7QMJ67C","created_at":"2026-07-05T05:36:51.155815+00:00"},{"alias_kind":"pith_short_8","alias_value":"PHDLNRYI","created_at":"2026-07-05T05:36:51.155815+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.08803","citing_title":"Some Essential Constructive Foundations for Systems and Control","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01968","citing_title":"AdamO: A Collapse-Suppressed Optimizer for Offline RL","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ","json":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ.json","graph_json":"https://pith.science/api/pith-number/PHDLNRYIW7QMJ67CQGWD74P3XZ/graph.json","events_json":"https://pith.science/api/pith-number/PHDLNRYIW7QMJ67CQGWD74P3XZ/events.json","paper":"https://pith.science/paper/PHDLNRYI"},"agent_actions":{"view_html":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ","download_json":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ.json","view_paper":"https://pith.science/paper/PHDLNRYI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.11092&json=true","fetch_graph":"https://pith.science/api/pith-number/PHDLNRYIW7QMJ67CQGWD74P3XZ/graph.json","fetch_events":"https://pith.science/api/pith-number/PHDLNRYIW7QMJ67CQGWD74P3XZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ/action/storage_attestation","attest_author":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ/action/author_attestation","sign_citation":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ/action/citation_signature","submit_replication":"https://pith.science/pith/PHDLNRYIW7QMJ67CQGWD74P3XZ/action/replication_record"}},"created_at":"2026-07-05T05:36:51.155815+00:00","updated_at":"2026-07-05T05:36:51.155815+00:00"}