{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:6XQIURETHVM7LHIVZXSPSDF4MS","short_pith_number":"pith:6XQIURET","schema_version":"1.0","canonical_sha256":"f5e08a44933d59f59d15cde4f90cbc64ad125c4c9f08a6c64019a2a03412378b","source":{"kind":"arxiv","id":"2309.15178","version":3},"attestation_state":"computed","paper":{"title":"Zero-Shot Reinforcement Learning from Low Quality Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jonathan M. Cullen, Scott Jeen, Tom Bewley","submitted_at":"2023-09-26T18:20:20Z","abstract_excerpt":"Zero-shot reinforcement learning (RL) promises to provide agents that can perform any task in an environment after an offline, reward-free pre-training phase. Methods leveraging successor measures and successor features have shown strong performance in this setting, but require access to large heterogenous datasets for pre-training which cannot be expected for most real problems. Here, we explore how the performance of zero-shot RL methods degrades when trained on small homogeneous datasets, and propose fixes inspired by conservatism, a well-established feature of performant single-task offlin"},"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":"2309.15178","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-09-26T18:20:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f429b66608f220048686f767f0347e0d179bb40cd7b44c2979c1ce2595ef79fe","abstract_canon_sha256":"cd43293d5c28e0a7165ee504526126b6b0dab1c67153e2c6f2f45c611dc9b4e9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:28:27.684365Z","signature_b64":"uuGZu5JUKnL8QTnlIMNHo/z6LVBfBR5OkI7zK1G2HWs2SGs9LHkfM1f2lo6YksDNduUuM8o1RP6MHGq+bUhqAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f5e08a44933d59f59d15cde4f90cbc64ad125c4c9f08a6c64019a2a03412378b","last_reissued_at":"2026-07-05T09:28:27.683857Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:28:27.683857Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Zero-Shot Reinforcement Learning from Low Quality Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jonathan M. Cullen, Scott Jeen, Tom Bewley","submitted_at":"2023-09-26T18:20:20Z","abstract_excerpt":"Zero-shot reinforcement learning (RL) promises to provide agents that can perform any task in an environment after an offline, reward-free pre-training phase. Methods leveraging successor measures and successor features have shown strong performance in this setting, but require access to large heterogenous datasets for pre-training which cannot be expected for most real problems. Here, we explore how the performance of zero-shot RL methods degrades when trained on small homogeneous datasets, and propose fixes inspired by conservatism, a well-established feature of performant single-task offlin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.15178","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/2309.15178/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":"2309.15178","created_at":"2026-07-05T09:28:27.683920+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.15178v3","created_at":"2026-07-05T09:28:27.683920+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.15178","created_at":"2026-07-05T09:28:27.683920+00:00"},{"alias_kind":"pith_short_12","alias_value":"6XQIURETHVM7","created_at":"2026-07-05T09:28:27.683920+00:00"},{"alias_kind":"pith_short_16","alias_value":"6XQIURETHVM7LHIV","created_at":"2026-07-05T09:28:27.683920+00:00"},{"alias_kind":"pith_short_8","alias_value":"6XQIURET","created_at":"2026-07-05T09:28:27.683920+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.05477","citing_title":"Epistemically-guided forward-backward exploration","ref_index":2024,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS","json":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS.json","graph_json":"https://pith.science/api/pith-number/6XQIURETHVM7LHIVZXSPSDF4MS/graph.json","events_json":"https://pith.science/api/pith-number/6XQIURETHVM7LHIVZXSPSDF4MS/events.json","paper":"https://pith.science/paper/6XQIURET"},"agent_actions":{"view_html":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS","download_json":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS.json","view_paper":"https://pith.science/paper/6XQIURET","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.15178&json=true","fetch_graph":"https://pith.science/api/pith-number/6XQIURETHVM7LHIVZXSPSDF4MS/graph.json","fetch_events":"https://pith.science/api/pith-number/6XQIURETHVM7LHIVZXSPSDF4MS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS/action/storage_attestation","attest_author":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS/action/author_attestation","sign_citation":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS/action/citation_signature","submit_replication":"https://pith.science/pith/6XQIURETHVM7LHIVZXSPSDF4MS/action/replication_record"}},"created_at":"2026-07-05T09:28:27.683920+00:00","updated_at":"2026-07-05T09:28:27.683920+00:00"}