{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:A7XVV5MNREPWEEBENR6SIGDVDE","short_pith_number":"pith:A7XVV5MN","schema_version":"1.0","canonical_sha256":"07ef5af58d891f6210246c7d241875190a39605b8860d66cd160a0cf9e569857","source":{"kind":"arxiv","id":"1802.09477","version":3},"attestation_state":"computed","paper":{"title":"Addressing Function Approximation Error in Actor-Critic Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.AI","authors_text":"David Meger, Herke van Hoof, Scott Fujimoto","submitted_at":"2018-02-26T17:54:49Z","abstract_excerpt":"In value-based reinforcement learning methods such as deep Q-learning, function approximation errors are known to lead to overestimated value estimates and suboptimal policies. We show that this problem persists in an actor-critic setting and propose novel mechanisms to minimize its effects on both the actor and the critic. Our algorithm builds on Double Q-learning, by taking the minimum value between a pair of critics to limit overestimation. We draw the connection between target networks and overestimation bias, and suggest delaying policy updates to reduce per-update error and further impro"},"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":"1802.09477","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2018-02-26T17:54:49Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"4f79674a0374b3a99bbf08a21a9cc1a9988de5a28a63721151c87e960682f02e","abstract_canon_sha256":"9fb330787100c9542d30ff1b9d88d59acb4d22c9808bdc2a9fe070ed60be2f5f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:02:45.178978Z","signature_b64":"0nC1hskNRuJ1N3zAdOIqEKGMzzd3/8DTvLJBxnGyJo9taRTdB+x9fL0xYHPQl9lkeYRSnMFvi3DZ/mmo4VMyDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"07ef5af58d891f6210246c7d241875190a39605b8860d66cd160a0cf9e569857","last_reissued_at":"2026-05-18T00:02:45.178554Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:02:45.178554Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Addressing Function Approximation Error in Actor-Critic Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.AI","authors_text":"David Meger, Herke van Hoof, Scott Fujimoto","submitted_at":"2018-02-26T17:54:49Z","abstract_excerpt":"In value-based reinforcement learning methods such as deep Q-learning, function approximation errors are known to lead to overestimated value estimates and suboptimal policies. We show that this problem persists in an actor-critic setting and propose novel mechanisms to minimize its effects on both the actor and the critic. Our algorithm builds on Double Q-learning, by taking the minimum value between a pair of critics to limit overestimation. We draw the connection between target networks and overestimation bias, and suggest delaying policy updates to reduce per-update error and further impro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1802.09477","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":""},"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":"1802.09477","created_at":"2026-05-18T00:02:45.178614+00:00"},{"alias_kind":"arxiv_version","alias_value":"1802.09477v3","created_at":"2026-05-18T00:02:45.178614+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1802.09477","created_at":"2026-05-18T00:02:45.178614+00:00"},{"alias_kind":"pith_short_12","alias_value":"A7XVV5MNREPW","created_at":"2026-05-18T12:32:13.499390+00:00"},{"alias_kind":"pith_short_16","alias_value":"A7XVV5MNREPWEEBE","created_at":"2026-05-18T12:32:13.499390+00:00"},{"alias_kind":"pith_short_8","alias_value":"A7XVV5MN","created_at":"2026-05-18T12:32:13.499390+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"1906.08649","citing_title":"Exploring Model-based Planning with Policy Networks","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"1907.02057","citing_title":"Benchmarking Model-Based Reinforcement Learning","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2604.13505","citing_title":"Cascaded TD3-PID Hybrid Controller for Quadrotor Trajectory Tracking in Wind Disturbance Environments","ref_index":33,"is_internal_anchor":true},{"citing_arxiv_id":"1812.05905","citing_title":"Soft Actor-Critic Algorithms and Applications","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"1801.01290","citing_title":"Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23073","citing_title":"RL Token: Bootstrapping Online RL with Vision-Language-Action Models","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11463","citing_title":"To Learn or Not to Learn: A Litmus Test for Using Reinforcement Learning in Control","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13505","citing_title":"Cascaded TD3-PID Hybrid Controller for Quadrotor Trajectory Tracking in Wind Disturbance Environments","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18190","citing_title":"Scalable Neighborhood-Based Multi-Agent Actor-Critic","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02552","citing_title":"Recurrent Deep Reinforcement Learning for Chemotherapy Control under Partial Observability","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE","json":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE.json","graph_json":"https://pith.science/api/pith-number/A7XVV5MNREPWEEBENR6SIGDVDE/graph.json","events_json":"https://pith.science/api/pith-number/A7XVV5MNREPWEEBENR6SIGDVDE/events.json","paper":"https://pith.science/paper/A7XVV5MN"},"agent_actions":{"view_html":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE","download_json":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE.json","view_paper":"https://pith.science/paper/A7XVV5MN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1802.09477&json=true","fetch_graph":"https://pith.science/api/pith-number/A7XVV5MNREPWEEBENR6SIGDVDE/graph.json","fetch_events":"https://pith.science/api/pith-number/A7XVV5MNREPWEEBENR6SIGDVDE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE/action/storage_attestation","attest_author":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE/action/author_attestation","sign_citation":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE/action/citation_signature","submit_replication":"https://pith.science/pith/A7XVV5MNREPWEEBENR6SIGDVDE/action/replication_record"}},"created_at":"2026-05-18T00:02:45.178614+00:00","updated_at":"2026-05-18T00:02:45.178614+00:00"}