{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:EE6FT336DFRNW4MKQ7IMN4GYXY","short_pith_number":"pith:EE6FT336","schema_version":"1.0","canonical_sha256":"213c59ef7e1962db718a87d0c6f0d8be27bd9019da418d8d3be5ddf3d3608dfd","source":{"kind":"arxiv","id":"2310.14550","version":3},"attestation_state":"computed","paper":{"title":"Corruption-Robust Offline Reinforcement Learning with General Function Approximation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chenlu Ye, Quanquan Gu, Rui Yang, Tong Zhang","submitted_at":"2023-10-23T04:07:26Z","abstract_excerpt":"We investigate the problem of corruption robustness in offline reinforcement learning (RL) with general function approximation, where an adversary can corrupt each sample in the offline dataset, and the corruption level $\\zeta\\geq0$ quantifies the cumulative corruption amount over $n$ episodes and $H$ steps. Our goal is to find a policy that is robust to such corruption and minimizes the suboptimality gap with respect to the optimal policy for the uncorrupted Markov decision processes (MDPs). Drawing inspiration from the uncertainty-weighting technique from the robust online RL setting \\citep{"},"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":"2310.14550","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-10-23T04:07:26Z","cross_cats_sorted":[],"title_canon_sha256":"fb53a0c3de21dab911f8d9250717209a8890aae334bf8bcd0d094c67af95a965","abstract_canon_sha256":"4fb7e1081295f80b285251ad83ff290418933b30fc0d14b7ef68f04fb4079add"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:46:25.208060Z","signature_b64":"osFQGJ5voIez9rAEcoUTGZYZF2iMcwyxTv7arBQFQgaeRNxdbgtU5qVypLNznGL4DwsP/4HIkDcQ0Qme4SX9Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"213c59ef7e1962db718a87d0c6f0d8be27bd9019da418d8d3be5ddf3d3608dfd","last_reissued_at":"2026-07-05T07:46:25.207633Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:46:25.207633Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Corruption-Robust Offline Reinforcement Learning with General Function Approximation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chenlu Ye, Quanquan Gu, Rui Yang, Tong Zhang","submitted_at":"2023-10-23T04:07:26Z","abstract_excerpt":"We investigate the problem of corruption robustness in offline reinforcement learning (RL) with general function approximation, where an adversary can corrupt each sample in the offline dataset, and the corruption level $\\zeta\\geq0$ quantifies the cumulative corruption amount over $n$ episodes and $H$ steps. Our goal is to find a policy that is robust to such corruption and minimizes the suboptimality gap with respect to the optimal policy for the uncorrupted Markov decision processes (MDPs). Drawing inspiration from the uncertainty-weighting technique from the robust online RL setting \\citep{"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.14550","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/2310.14550/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":"2310.14550","created_at":"2026-07-05T07:46:25.207690+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.14550v3","created_at":"2026-07-05T07:46:25.207690+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.14550","created_at":"2026-07-05T07:46:25.207690+00:00"},{"alias_kind":"pith_short_12","alias_value":"EE6FT336DFRN","created_at":"2026-07-05T07:46:25.207690+00:00"},{"alias_kind":"pith_short_16","alias_value":"EE6FT336DFRNW4MK","created_at":"2026-07-05T07:46:25.207690+00:00"},{"alias_kind":"pith_short_8","alias_value":"EE6FT336","created_at":"2026-07-05T07:46:25.207690+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/EE6FT336DFRNW4MKQ7IMN4GYXY","json":"https://pith.science/pith/EE6FT336DFRNW4MKQ7IMN4GYXY.json","graph_json":"https://pith.science/api/pith-number/EE6FT336DFRNW4MKQ7IMN4GYXY/graph.json","events_json":"https://pith.science/api/pith-number/EE6FT336DFRNW4MKQ7IMN4GYXY/events.json","paper":"https://pith.science/paper/EE6FT336"},"agent_actions":{"view_html":"https://pith.science/pith/EE6FT336DFRNW4MKQ7IMN4GYXY","download_json":"https://pith.science/pith/EE6FT336DFRNW4MKQ7IMN4GYXY.json","view_paper":"https://pith.science/paper/EE6FT336","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.14550&json=true","fetch_graph":"https://pith.science/api/pith-number/EE6FT336DFRNW4MKQ7IMN4GYXY/graph.json","fetch_events":"https://pith.science/api/pith-number/EE6FT336DFRNW4MKQ7IMN4GYXY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EE6FT336DFRNW4MKQ7IMN4GYXY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EE6FT336DFRNW4MKQ7IMN4GYXY/action/storage_attestation","attest_author":"https://pith.science/pith/EE6FT336DFRNW4MKQ7IMN4GYXY/action/author_attestation","sign_citation":"https://pith.science/pith/EE6FT336DFRNW4MKQ7IMN4GYXY/action/citation_signature","submit_replication":"https://pith.science/pith/EE6FT336DFRNW4MKQ7IMN4GYXY/action/replication_record"}},"created_at":"2026-07-05T07:46:25.207690+00:00","updated_at":"2026-07-05T07:46:25.207690+00:00"}