{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2015:DKREUUR6KM7OB22JL2S7KLQAPI","short_pith_number":"pith:DKREUUR6","schema_version":"1.0","canonical_sha256":"1aa24a523e533ee0eb495ea5f52e007a09f9e005c3a3940c108ca5933da3f641","source":{"kind":"arxiv","id":"1511.06295","version":2},"attestation_state":"computed","paper":{"title":"Policy Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andrei A. Rusu, Caglar Gulcehre, Guillaume Desjardins, James Kirkpatrick, Koray Kavukcuoglu, Raia Hadsell, Razvan Pascanu, Sergio Gomez Colmenarejo, Volodymyr Mnih","submitted_at":"2015-11-19T18:38:47Z","abstract_excerpt":"Policies for complex visual tasks have been successfully learned with deep reinforcement learning, using an approach called deep Q-networks (DQN), but relatively large (task-specific) networks and extensive training are needed to achieve good performance. In this work, we present a novel method called policy distillation that can be used to extract the policy of a reinforcement learning agent and train a new network that performs at the expert level while being dramatically smaller and more efficient. Furthermore, the same method can be used to consolidate multiple task-specific policies into "},"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":"1511.06295","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2015-11-19T18:38:47Z","cross_cats_sorted":[],"title_canon_sha256":"d3db49106b23606ca82e4a74cbf86ba342f17b44d61f5c0f4d5fedb3452fd071","abstract_canon_sha256":"ecbb6b1efcbeb6ef31dfc0f75d3dc61a18bfaaf2351787b17758d35f61cc3fc3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:23:12.982881Z","signature_b64":"HGZUykh7ZwbXTHYXAYzNehyqOPNu7vaqgMLa2JRTuy0fVHhuzqUMrSf2+vQAS4B+4pX7gjY5Jmyw/gcuJMkUDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1aa24a523e533ee0eb495ea5f52e007a09f9e005c3a3940c108ca5933da3f641","last_reissued_at":"2026-05-18T01:23:12.982188Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:23:12.982188Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Policy Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andrei A. Rusu, Caglar Gulcehre, Guillaume Desjardins, James Kirkpatrick, Koray Kavukcuoglu, Raia Hadsell, Razvan Pascanu, Sergio Gomez Colmenarejo, Volodymyr Mnih","submitted_at":"2015-11-19T18:38:47Z","abstract_excerpt":"Policies for complex visual tasks have been successfully learned with deep reinforcement learning, using an approach called deep Q-networks (DQN), but relatively large (task-specific) networks and extensive training are needed to achieve good performance. In this work, we present a novel method called policy distillation that can be used to extract the policy of a reinforcement learning agent and train a new network that performs at the expert level while being dramatically smaller and more efficient. Furthermore, the same method can be used to consolidate multiple task-specific policies into "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1511.06295","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":""},"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":"1511.06295","created_at":"2026-05-18T01:23:12.982323+00:00"},{"alias_kind":"arxiv_version","alias_value":"1511.06295v2","created_at":"2026-05-18T01:23:12.982323+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1511.06295","created_at":"2026-05-18T01:23:12.982323+00:00"},{"alias_kind":"pith_short_12","alias_value":"DKREUUR6KM7O","created_at":"2026-05-18T12:29:17.054201+00:00"},{"alias_kind":"pith_short_16","alias_value":"DKREUUR6KM7OB22J","created_at":"2026-05-18T12:29:17.054201+00:00"},{"alias_kind":"pith_short_8","alias_value":"DKREUUR6","created_at":"2026-05-18T12:29:17.054201+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":17,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2605.14022","citing_title":"Policy-DRIFT: Dynamic Reward-Informed Flow Trajectory Steering","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13665","citing_title":"Robot Squid Game: Quadrupedal Locomotion for Traversing Narrow Tunnels","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10648","citing_title":"Demystifying Deep Reinforcement Learning: A Neuro-Symbolic Framework for Interpretable Open RAN Automation","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2306.08543","citing_title":"MiniLLM: On-Policy Distillation of Large Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"1606.04671","citing_title":"Progressive Neural Networks","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27411","citing_title":"Detecting is Easy, Adapting is Hard: Local Expert Growth for Visual Model-Based Reinforcement Learning under Distribution Shift","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06094","citing_title":"VISD: Enhancing Video Reasoning via Structured Self-Distillation","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10648","citing_title":"Demystifying Deep Reinforcement Learning: A Neuro-Symbolic Framework for Interpretable Open RAN Automation","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06094","citing_title":"VISD: Enhancing Video Reasoning via Structured Self-Distillation","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05478","citing_title":"LANTERN: LLM-Augmented Neurosymbolic Transfer with Experience-Gated Reasoning Networks","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08527","citing_title":"Demystifying OPD: Length Inflation and Stabilization Strategies for Large Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09452","citing_title":"SafeAdapt: Provably Safe Policy Updates in Deep Reinforcement Learning","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06094","citing_title":"VISD: Enhancing Video Reasoning via Structured Self-Distillation","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05828","citing_title":"Precise Aggressive Aerial Maneuvers with Sensorimotor Policies","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14360","citing_title":"Digital Guardians: The Past and The Future of Cyber-Physical Resilience","ref_index":196,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17800","citing_title":"ReFineVLA: Multimodal Reasoning-Aware Generalist Robotic Policies via Teacher-Guided Fine-Tuning","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02159","citing_title":"Combining Trained Models in Reinforcement Learning","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI","json":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI.json","graph_json":"https://pith.science/api/pith-number/DKREUUR6KM7OB22JL2S7KLQAPI/graph.json","events_json":"https://pith.science/api/pith-number/DKREUUR6KM7OB22JL2S7KLQAPI/events.json","paper":"https://pith.science/paper/DKREUUR6"},"agent_actions":{"view_html":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI","download_json":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI.json","view_paper":"https://pith.science/paper/DKREUUR6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1511.06295&json=true","fetch_graph":"https://pith.science/api/pith-number/DKREUUR6KM7OB22JL2S7KLQAPI/graph.json","fetch_events":"https://pith.science/api/pith-number/DKREUUR6KM7OB22JL2S7KLQAPI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI/action/storage_attestation","attest_author":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI/action/author_attestation","sign_citation":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI/action/citation_signature","submit_replication":"https://pith.science/pith/DKREUUR6KM7OB22JL2S7KLQAPI/action/replication_record"}},"created_at":"2026-05-18T01:23:12.982323+00:00","updated_at":"2026-05-18T01:23:12.982323+00:00"}