{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:FP4IWC3F7BGFVLI6Q2OOU6ANZD","short_pith_number":"pith:FP4IWC3F","schema_version":"1.0","canonical_sha256":"2bf88b0b65f84c5aad1e869cea780dc8f6353ff19386348647b0f32688c45f15","source":{"kind":"arxiv","id":"2102.13515","version":3},"attestation_state":"computed","paper":{"title":"Beyond Fine-Tuning: Transferring Behavior in Reinforcement Learning","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Adri\\`a Puigdom\\`enech Badia, Alex Vitvitskyi, Andre Barreto, Charles Blundell, Pablo Sprechmann, Steven Hansen, Steven Kapturowski, V\\'ictor Campos","submitted_at":"2021-02-24T16:51:02Z","abstract_excerpt":"Designing agents that acquire knowledge autonomously and use it to solve new tasks efficiently is an important challenge in reinforcement learning. Knowledge acquired during an unsupervised pre-training phase is often transferred by fine-tuning neural network weights once rewards are exposed, as is common practice in supervised domains. Given the nature of the reinforcement learning problem, we argue that standard fine-tuning strategies alone are not enough for efficient transfer in challenging domains. We introduce Behavior Transfer (BT), a technique that leverages pre-trained policies for ex"},"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":"2102.13515","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2021-02-24T16:51:02Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"00210d87d1d4ea19880cad732199ca36f7c91d68f024dc92de6ade84fd6f1a58","abstract_canon_sha256":"f3f9fa27432e3a1d4ff8e94d026ff8e3f3518e5237291a2b546d0c5663c960b6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:47:05.274298Z","signature_b64":"VoUpPb4BTYM9tM3qWVxwfTxmL7f4lXbq52mVe54YZumw7V8voTI2YieEIirv9yTtbPuflxVoJujYXsujxwiFDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2bf88b0b65f84c5aad1e869cea780dc8f6353ff19386348647b0f32688c45f15","last_reissued_at":"2026-07-05T02:47:05.273388Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:47:05.273388Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Fine-Tuning: Transferring Behavior in Reinforcement Learning","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Adri\\`a Puigdom\\`enech Badia, Alex Vitvitskyi, Andre Barreto, Charles Blundell, Pablo Sprechmann, Steven Hansen, Steven Kapturowski, V\\'ictor Campos","submitted_at":"2021-02-24T16:51:02Z","abstract_excerpt":"Designing agents that acquire knowledge autonomously and use it to solve new tasks efficiently is an important challenge in reinforcement learning. Knowledge acquired during an unsupervised pre-training phase is often transferred by fine-tuning neural network weights once rewards are exposed, as is common practice in supervised domains. Given the nature of the reinforcement learning problem, we argue that standard fine-tuning strategies alone are not enough for efficient transfer in challenging domains. We introduce Behavior Transfer (BT), a technique that leverages pre-trained policies for ex"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.13515","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/2102.13515/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":"2102.13515","created_at":"2026-07-05T02:47:05.273470+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.13515v3","created_at":"2026-07-05T02:47:05.273470+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.13515","created_at":"2026-07-05T02:47:05.273470+00:00"},{"alias_kind":"pith_short_12","alias_value":"FP4IWC3F7BGF","created_at":"2026-07-05T02:47:05.273470+00:00"},{"alias_kind":"pith_short_16","alias_value":"FP4IWC3F7BGFVLI6","created_at":"2026-07-05T02:47:05.273470+00:00"},{"alias_kind":"pith_short_8","alias_value":"FP4IWC3F","created_at":"2026-07-05T02:47:05.273470+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.00913","citing_title":"Fine-Tuning without Performance Degradation","ref_index":4,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD","json":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD.json","graph_json":"https://pith.science/api/pith-number/FP4IWC3F7BGFVLI6Q2OOU6ANZD/graph.json","events_json":"https://pith.science/api/pith-number/FP4IWC3F7BGFVLI6Q2OOU6ANZD/events.json","paper":"https://pith.science/paper/FP4IWC3F"},"agent_actions":{"view_html":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD","download_json":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD.json","view_paper":"https://pith.science/paper/FP4IWC3F","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.13515&json=true","fetch_graph":"https://pith.science/api/pith-number/FP4IWC3F7BGFVLI6Q2OOU6ANZD/graph.json","fetch_events":"https://pith.science/api/pith-number/FP4IWC3F7BGFVLI6Q2OOU6ANZD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD/action/storage_attestation","attest_author":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD/action/author_attestation","sign_citation":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD/action/citation_signature","submit_replication":"https://pith.science/pith/FP4IWC3F7BGFVLI6Q2OOU6ANZD/action/replication_record"}},"created_at":"2026-07-05T02:47:05.273470+00:00","updated_at":"2026-07-05T02:47:05.273470+00:00"}