{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:C36ZRCPOE2X6HKV7HCFXZNBX2B","short_pith_number":"pith:C36ZRCPO","schema_version":"1.0","canonical_sha256":"16fd9889ee26afe3aabf388b7cb437d06a197a3d5a8557ca2b62fa6f75eeab32","source":{"kind":"arxiv","id":"2109.14419","version":3},"attestation_state":"computed","paper":{"title":"On the Estimation Bias in Double Q-Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Beining Han, Chongjie Zhang, Guangxiang Zhu, Hao Hu, Jianglun Chen, Zhizhou Ren","submitted_at":"2021-09-29T13:41:24Z","abstract_excerpt":"Double Q-learning is a classical method for reducing overestimation bias, which is caused by taking maximum estimated values in the Bellman operation. Its variants in the deep Q-learning paradigm have shown great promise in producing reliable value prediction and improving learning performance. However, as shown by prior work, double Q-learning is not fully unbiased and suffers from underestimation bias. In this paper, we show that such underestimation bias may lead to multiple non-optimal fixed points under an approximate Bellman operator. To address the concerns of converging to non-optimal "},"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":"2109.14419","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-09-29T13:41:24Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"3bf8f32b2425f2e4598051a9acec6e61ad615d4184839e3b184d572b850ab2a7","abstract_canon_sha256":"327fa02347c6e1b376d1f17b026f6cc7ff6d003c349b675441d3969e59afd489"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:48:19.799348Z","signature_b64":"G99li88Dn7pS0rMTavsR7Y1zNIxWEm5kd+FW0S2UkclsZSQCAcCB8oDA+wSfMa47N40tczEDiwOZ/pdvGb+MDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"16fd9889ee26afe3aabf388b7cb437d06a197a3d5a8557ca2b62fa6f75eeab32","last_reissued_at":"2026-07-05T03:48:19.798908Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:48:19.798908Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Estimation Bias in Double Q-Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Beining Han, Chongjie Zhang, Guangxiang Zhu, Hao Hu, Jianglun Chen, Zhizhou Ren","submitted_at":"2021-09-29T13:41:24Z","abstract_excerpt":"Double Q-learning is a classical method for reducing overestimation bias, which is caused by taking maximum estimated values in the Bellman operation. Its variants in the deep Q-learning paradigm have shown great promise in producing reliable value prediction and improving learning performance. However, as shown by prior work, double Q-learning is not fully unbiased and suffers from underestimation bias. In this paper, we show that such underestimation bias may lead to multiple non-optimal fixed points under an approximate Bellman operator. To address the concerns of converging to non-optimal "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2109.14419","kind":"arxiv","version":3},"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/2109.14419/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":"2109.14419","created_at":"2026-07-05T03:48:19.798964+00:00"},{"alias_kind":"arxiv_version","alias_value":"2109.14419v3","created_at":"2026-07-05T03:48:19.798964+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2109.14419","created_at":"2026-07-05T03:48:19.798964+00:00"},{"alias_kind":"pith_short_12","alias_value":"C36ZRCPOE2X6","created_at":"2026-07-05T03:48:19.798964+00:00"},{"alias_kind":"pith_short_16","alias_value":"C36ZRCPOE2X6HKV7","created_at":"2026-07-05T03:48:19.798964+00:00"},{"alias_kind":"pith_short_8","alias_value":"C36ZRCPO","created_at":"2026-07-05T03:48:19.798964+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/C36ZRCPOE2X6HKV7HCFXZNBX2B","json":"https://pith.science/pith/C36ZRCPOE2X6HKV7HCFXZNBX2B.json","graph_json":"https://pith.science/api/pith-number/C36ZRCPOE2X6HKV7HCFXZNBX2B/graph.json","events_json":"https://pith.science/api/pith-number/C36ZRCPOE2X6HKV7HCFXZNBX2B/events.json","paper":"https://pith.science/paper/C36ZRCPO"},"agent_actions":{"view_html":"https://pith.science/pith/C36ZRCPOE2X6HKV7HCFXZNBX2B","download_json":"https://pith.science/pith/C36ZRCPOE2X6HKV7HCFXZNBX2B.json","view_paper":"https://pith.science/paper/C36ZRCPO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2109.14419&json=true","fetch_graph":"https://pith.science/api/pith-number/C36ZRCPOE2X6HKV7HCFXZNBX2B/graph.json","fetch_events":"https://pith.science/api/pith-number/C36ZRCPOE2X6HKV7HCFXZNBX2B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C36ZRCPOE2X6HKV7HCFXZNBX2B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C36ZRCPOE2X6HKV7HCFXZNBX2B/action/storage_attestation","attest_author":"https://pith.science/pith/C36ZRCPOE2X6HKV7HCFXZNBX2B/action/author_attestation","sign_citation":"https://pith.science/pith/C36ZRCPOE2X6HKV7HCFXZNBX2B/action/citation_signature","submit_replication":"https://pith.science/pith/C36ZRCPOE2X6HKV7HCFXZNBX2B/action/replication_record"}},"created_at":"2026-07-05T03:48:19.798964+00:00","updated_at":"2026-07-05T03:48:19.798964+00:00"}