{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:UOXCMSN77E6ZKZOEMVC7B6IJXZ","short_pith_number":"pith:UOXCMSN7","schema_version":"1.0","canonical_sha256":"a3ae2649bff93d9565c46545f0f909be45943d1f6992fcbbbee77a454f9edd85","source":{"kind":"arxiv","id":"2010.13611","version":3},"attestation_state":"computed","paper":{"title":"OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anurag Ajay, Aviral Kumar, Ofir Nachum, Pulkit Agrawal, Sergey Levine","submitted_at":"2020-10-26T14:31:08Z","abstract_excerpt":"Reinforcement learning (RL) has achieved impressive performance in a variety of online settings in which an agent's ability to query the environment for transitions and rewards is effectively unlimited. However, in many practical applications, the situation is reversed: an agent may have access to large amounts of undirected offline experience data, while access to the online environment is severely limited. In this work, we focus on this offline setting. Our main insight is that, when presented with offline data composed of a variety of behaviors, an effective way to leverage this data is to "},"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":"2010.13611","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-10-26T14:31:08Z","cross_cats_sorted":[],"title_canon_sha256":"b97045fe3432548a4037281c236e8855c82ad4837e686986d14f75ac5034ef03","abstract_canon_sha256":"a7f5ff0bb2e8ba3e8b7ab4269f4393561bb8804424a89d160b807f3e00ea6df4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:37:42.555989Z","signature_b64":"scHyzzcfK4IvIbCsarnEAlvG13VczITJOwHR+pdBf2HbFb03rRXDGmnt8hojS/qluxgX5IUj9/+fO9lQJNCIBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a3ae2649bff93d9565c46545f0f909be45943d1f6992fcbbbee77a454f9edd85","last_reissued_at":"2026-07-05T02:37:42.555481Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:37:42.555481Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anurag Ajay, Aviral Kumar, Ofir Nachum, Pulkit Agrawal, Sergey Levine","submitted_at":"2020-10-26T14:31:08Z","abstract_excerpt":"Reinforcement learning (RL) has achieved impressive performance in a variety of online settings in which an agent's ability to query the environment for transitions and rewards is effectively unlimited. However, in many practical applications, the situation is reversed: an agent may have access to large amounts of undirected offline experience data, while access to the online environment is severely limited. In this work, we focus on this offline setting. Our main insight is that, when presented with offline data composed of a variety of behaviors, an effective way to leverage this data is to "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.13611","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/2010.13611/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":"2010.13611","created_at":"2026-07-05T02:37:42.555545+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.13611v3","created_at":"2026-07-05T02:37:42.555545+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.13611","created_at":"2026-07-05T02:37:42.555545+00:00"},{"alias_kind":"pith_short_12","alias_value":"UOXCMSN77E6Z","created_at":"2026-07-05T02:37:42.555545+00:00"},{"alias_kind":"pith_short_16","alias_value":"UOXCMSN77E6ZKZOE","created_at":"2026-07-05T02:37:42.555545+00:00"},{"alias_kind":"pith_short_8","alias_value":"UOXCMSN7","created_at":"2026-07-05T02:37:42.555545+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09115","citing_title":"Counterfactual Transport Flows for Offline Conservative Trajectory Refinement","ref_index":81,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31958","citing_title":"Adapting Generalist Robot Policies with Semantic Reinforcement Learning","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29201","citing_title":"Behavior Uncloning: Distilling Mode Redirection into Policy Weights without Inference-Time Steering","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21822","citing_title":"Implicit Safety Alignment from Crowd Preferences","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2106.01345","citing_title":"Decision Transformer: Reinforcement Learning via Sequence Modeling","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2108.03298","citing_title":"What Matters in Learning from Offline Human Demonstrations for Robot Manipulation","ref_index":48,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ","json":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ.json","graph_json":"https://pith.science/api/pith-number/UOXCMSN77E6ZKZOEMVC7B6IJXZ/graph.json","events_json":"https://pith.science/api/pith-number/UOXCMSN77E6ZKZOEMVC7B6IJXZ/events.json","paper":"https://pith.science/paper/UOXCMSN7"},"agent_actions":{"view_html":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ","download_json":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ.json","view_paper":"https://pith.science/paper/UOXCMSN7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.13611&json=true","fetch_graph":"https://pith.science/api/pith-number/UOXCMSN77E6ZKZOEMVC7B6IJXZ/graph.json","fetch_events":"https://pith.science/api/pith-number/UOXCMSN77E6ZKZOEMVC7B6IJXZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ/action/storage_attestation","attest_author":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ/action/author_attestation","sign_citation":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ/action/citation_signature","submit_replication":"https://pith.science/pith/UOXCMSN77E6ZKZOEMVC7B6IJXZ/action/replication_record"}},"created_at":"2026-07-05T02:37:42.555545+00:00","updated_at":"2026-07-05T02:37:42.555545+00:00"}