{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:BYHOZFD6YEIUB5G7GWN2S62QUB","short_pith_number":"pith:BYHOZFD6","schema_version":"1.0","canonical_sha256":"0e0eec947ec11140f4df359ba97b50a07756b5d85d63794135507f696fc1eaa4","source":{"kind":"arxiv","id":"2202.12742","version":2},"attestation_state":"computed","paper":{"title":"Learning Relative Return Policies With Upside-Down Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Dylan R. Ashley, J\\\"urgen Schmidhuber, Kai Arulkumaran, Rupesh Kumar Srivastava","submitted_at":"2022-02-23T07:21:44Z","abstract_excerpt":"Lately, there has been a resurgence of interest in using supervised learning to solve reinforcement learning problems. Recent work in this area has largely focused on learning command-conditioned policies. We investigate the potential of one such method -- upside-down reinforcement learning -- to work with commands that specify a desired relationship between some scalar value and the observed return. We show that upside-down reinforcement learning can learn to carry out such commands online in a tabular bandit setting and in CartPole with non-linear function approximation. By doing so, we demo"},"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":"2202.12742","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-02-23T07:21:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9197f4066281e80c54817b291f2c9ad2bf8b1097f0ff6cad3c2b413c0ac35f04","abstract_canon_sha256":"a105343d919caec3b4459e6c30d764d42cf5cb510ca2419e5fb88c1cc1f38b52"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:22:04.206904Z","signature_b64":"Tnu/fQr/UNTJ4RUpJz5T5HNUsCUq2AnKQBzg+T9ByuR0Av/GrpUrQOsNL+s1paZFyHQTpbYkbP32BjElc71qBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0e0eec947ec11140f4df359ba97b50a07756b5d85d63794135507f696fc1eaa4","last_reissued_at":"2026-07-05T04:22:04.206462Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:22:04.206462Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Relative Return Policies With Upside-Down Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Dylan R. Ashley, J\\\"urgen Schmidhuber, Kai Arulkumaran, Rupesh Kumar Srivastava","submitted_at":"2022-02-23T07:21:44Z","abstract_excerpt":"Lately, there has been a resurgence of interest in using supervised learning to solve reinforcement learning problems. Recent work in this area has largely focused on learning command-conditioned policies. We investigate the potential of one such method -- upside-down reinforcement learning -- to work with commands that specify a desired relationship between some scalar value and the observed return. We show that upside-down reinforcement learning can learn to carry out such commands online in a tabular bandit setting and in CartPole with non-linear function approximation. By doing so, we demo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.12742","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/2202.12742/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":"2202.12742","created_at":"2026-07-05T04:22:04.206513+00:00"},{"alias_kind":"arxiv_version","alias_value":"2202.12742v2","created_at":"2026-07-05T04:22:04.206513+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.12742","created_at":"2026-07-05T04:22:04.206513+00:00"},{"alias_kind":"pith_short_12","alias_value":"BYHOZFD6YEIU","created_at":"2026-07-05T04:22:04.206513+00:00"},{"alias_kind":"pith_short_16","alias_value":"BYHOZFD6YEIUB5G7","created_at":"2026-07-05T04:22:04.206513+00:00"},{"alias_kind":"pith_short_8","alias_value":"BYHOZFD6","created_at":"2026-07-05T04:22:04.206513+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.11457","citing_title":"Upside-Down Reinforcement Learning for More Interpretable Optimal Control","ref_index":5,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB","json":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB.json","graph_json":"https://pith.science/api/pith-number/BYHOZFD6YEIUB5G7GWN2S62QUB/graph.json","events_json":"https://pith.science/api/pith-number/BYHOZFD6YEIUB5G7GWN2S62QUB/events.json","paper":"https://pith.science/paper/BYHOZFD6"},"agent_actions":{"view_html":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB","download_json":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB.json","view_paper":"https://pith.science/paper/BYHOZFD6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2202.12742&json=true","fetch_graph":"https://pith.science/api/pith-number/BYHOZFD6YEIUB5G7GWN2S62QUB/graph.json","fetch_events":"https://pith.science/api/pith-number/BYHOZFD6YEIUB5G7GWN2S62QUB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB/action/storage_attestation","attest_author":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB/action/author_attestation","sign_citation":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB/action/citation_signature","submit_replication":"https://pith.science/pith/BYHOZFD6YEIUB5G7GWN2S62QUB/action/replication_record"}},"created_at":"2026-07-05T04:22:04.206513+00:00","updated_at":"2026-07-05T04:22:04.206513+00:00"}