{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:2YNZJQXKERHE44UGTNN3W6KL3R","short_pith_number":"pith:2YNZJQXK","schema_version":"1.0","canonical_sha256":"d61b94c2ea244e4e72869b5bbb794bdc6a3c7c3d2353b502e82050d6f4510a50","source":{"kind":"arxiv","id":"2603.05066","version":3},"attestation_state":"computed","paper":{"title":"Reward-Conditioned Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Conditioning RL agents on reward parameters during single-objective training enables zero-shot adaptation to new rewards via replay data alone.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Marek Cygan, Michal Nauman, Pieter Abbeel","submitted_at":"2026-03-05T11:29:17Z","abstract_excerpt":"Single-task RL agents are typically trained under a fixed reward function, which limits their robustness to reward misspecification and their ability to adapt to changing preferences. We introduce Reward-Conditioned Reinforcement Learning (RCRL), an off-policy method that conditions agents on reward parameterizations while collecting experience under a single nominal objective. By recomputing counterfactual rewards from shared replay data, RCRL exposes the agent to multiple reward objectives without additional environment interaction, connecting single-task RL with ideas from multi-objective a"},"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":true},"canonical_record":{"source":{"id":"2603.05066","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-03-05T11:29:17Z","cross_cats_sorted":[],"title_canon_sha256":"70cb4e591b6b3f80a2eb6f4194501f2e5ff9d1017e8ad8e834565b90df3e761d","abstract_canon_sha256":"c5b4a4bb45cbd5cc2bcec4f144608d501ff9ef6fb5eb15de46ef0744c9be4430"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:06:08.894232Z","signature_b64":"XU8XaRUzEfBrp0dERnYGULt1RzB5bG7ZSyGtADM9dPyD6ndM4/aaBMpeIddVSWhCyfbZxqhLao+0Q0b2lT7jBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d61b94c2ea244e4e72869b5bbb794bdc6a3c7c3d2353b502e82050d6f4510a50","last_reissued_at":"2026-05-20T01:06:08.893422Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:06:08.893422Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reward-Conditioned Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Conditioning RL agents on reward parameters during single-objective training enables zero-shot adaptation to new rewards via replay data alone.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Marek Cygan, Michal Nauman, Pieter Abbeel","submitted_at":"2026-03-05T11:29:17Z","abstract_excerpt":"Single-task RL agents are typically trained under a fixed reward function, which limits their robustness to reward misspecification and their ability to adapt to changing preferences. We introduce Reward-Conditioned Reinforcement Learning (RCRL), an off-policy method that conditions agents on reward parameterizations while collecting experience under a single nominal objective. By recomputing counterfactual rewards from shared replay data, RCRL exposes the agent to multiple reward objectives without additional environment interaction, connecting single-task RL with ideas from multi-objective a"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"RCRL improves sample efficiency under the nominal reward parameterization, enables efficient adaptation to new parameterizations, and supports zero-shot behavioral adjustment at deployment.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That recomputing counterfactual rewards from replay data collected under the nominal policy produces unbiased training signals for other reward parameterizations.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"RCRL conditions RL policies on reward parameters and uses shared replay data to train for multiple objectives under a single nominal reward, improving efficiency and adaptability.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Conditioning RL agents on reward parameters during single-objective training enables zero-shot adaptation to new rewards via replay data alone.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ec667cc40c61ec88d2bb4450fb4e3a40c40f419c5ecff6bbb8650ca19c74c784"},"source":{"id":"2603.05066","kind":"arxiv","version":3},"verdict":{"id":"abd32904-dc8c-45a7-ab96-799f4e4d18f8","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T16:05:50.794154Z","strongest_claim":"RCRL improves sample efficiency under the nominal reward parameterization, enables efficient adaptation to new parameterizations, and supports zero-shot behavioral adjustment at deployment.","one_line_summary":"RCRL conditions RL policies on reward parameters and uses shared replay data to train for multiple objectives under a single nominal reward, improving efficiency and adaptability.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That recomputing counterfactual rewards from replay data collected under the nominal policy produces unbiased training signals for other reward parameterizations.","pith_extraction_headline":"Conditioning RL agents on reward parameters during single-objective training enables zero-shot adaptation to new rewards via replay data alone."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.05066/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":2,"snapshot_sha256":"ef1afa9b6df446ae10a438a688321c081a7446c6e8340ec9c70cb8bf942e7a4e"},"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":"2603.05066","created_at":"2026-05-20T01:06:08.893534+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.05066v3","created_at":"2026-05-20T01:06:08.893534+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.05066","created_at":"2026-05-20T01:06:08.893534+00:00"},{"alias_kind":"pith_short_12","alias_value":"2YNZJQXKERHE","created_at":"2026-05-20T01:06:08.893534+00:00"},{"alias_kind":"pith_short_16","alias_value":"2YNZJQXKERHE44UG","created_at":"2026-05-20T01:06:08.893534+00:00"},{"alias_kind":"pith_short_8","alias_value":"2YNZJQXK","created_at":"2026-05-20T01:06:08.893534+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R","json":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R.json","graph_json":"https://pith.science/api/pith-number/2YNZJQXKERHE44UGTNN3W6KL3R/graph.json","events_json":"https://pith.science/api/pith-number/2YNZJQXKERHE44UGTNN3W6KL3R/events.json","paper":"https://pith.science/paper/2YNZJQXK"},"agent_actions":{"view_html":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R","download_json":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R.json","view_paper":"https://pith.science/paper/2YNZJQXK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.05066&json=true","fetch_graph":"https://pith.science/api/pith-number/2YNZJQXKERHE44UGTNN3W6KL3R/graph.json","fetch_events":"https://pith.science/api/pith-number/2YNZJQXKERHE44UGTNN3W6KL3R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R/action/storage_attestation","attest_author":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R/action/author_attestation","sign_citation":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R/action/citation_signature","submit_replication":"https://pith.science/pith/2YNZJQXKERHE44UGTNN3W6KL3R/action/replication_record"}},"created_at":"2026-05-20T01:06:08.893534+00:00","updated_at":"2026-05-20T01:06:08.893534+00:00"}