{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:RNGRIJL4HVGRUWOZ365I3MB65T","short_pith_number":"pith:RNGRIJL4","schema_version":"1.0","canonical_sha256":"8b4d14257c3d4d1a59d9dfba8db03eecc1e91f8e40ddef472d0df1c9dbdccb25","source":{"kind":"arxiv","id":"1703.03864","version":2},"attestation_state":"computed","paper":{"title":"Evolution Strategies as a Scalable Alternative to Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.NE"],"primary_cat":"stat.ML","authors_text":"Ilya Sutskever, Jonathan Ho, Szymon Sidor, Tim Salimans, Xi Chen","submitted_at":"2017-03-10T23:02:19Z","abstract_excerpt":"We explore the use of Evolution Strategies (ES), a class of black box optimization algorithms, as an alternative to popular MDP-based RL techniques such as Q-learning and Policy Gradients. Experiments on MuJoCo and Atari show that ES is a viable solution strategy that scales extremely well with the number of CPUs available: By using a novel communication strategy based on common random numbers, our ES implementation only needs to communicate scalars, making it possible to scale to over a thousand parallel workers. This allows us to solve 3D humanoid walking in 10 minutes and obtain competitive"},"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":"1703.03864","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2017-03-10T23:02:19Z","cross_cats_sorted":["cs.AI","cs.LG","cs.NE"],"title_canon_sha256":"621fdf04f6a7ecce24860b6a43cb7392371960e54a33f9a5668de90859aed9fc","abstract_canon_sha256":"b51868ea0ae6c040f9d4afc1ad404a73f9acf582cfbd0d4203eccf6a3ce75237"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:35:46.479990Z","signature_b64":"BBpTkO9IN82H2t1dSVWA3BrwIUO3gq3mZzz8CgzB7wGQlwJ7tBLHj6YBSVI6Q1LXJ/5A1yxdWetJ+XMcHqUxAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8b4d14257c3d4d1a59d9dfba8db03eecc1e91f8e40ddef472d0df1c9dbdccb25","last_reissued_at":"2026-05-18T00:35:46.479381Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:35:46.479381Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evolution Strategies as a Scalable Alternative to Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.NE"],"primary_cat":"stat.ML","authors_text":"Ilya Sutskever, Jonathan Ho, Szymon Sidor, Tim Salimans, Xi Chen","submitted_at":"2017-03-10T23:02:19Z","abstract_excerpt":"We explore the use of Evolution Strategies (ES), a class of black box optimization algorithms, as an alternative to popular MDP-based RL techniques such as Q-learning and Policy Gradients. Experiments on MuJoCo and Atari show that ES is a viable solution strategy that scales extremely well with the number of CPUs available: By using a novel communication strategy based on common random numbers, our ES implementation only needs to communicate scalars, making it possible to scale to over a thousand parallel workers. This allows us to solve 3D humanoid walking in 10 minutes and obtain competitive"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1703.03864","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":"1703.03864","created_at":"2026-05-18T00:35:46.479492+00:00"},{"alias_kind":"arxiv_version","alias_value":"1703.03864v2","created_at":"2026-05-18T00:35:46.479492+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1703.03864","created_at":"2026-05-18T00:35:46.479492+00:00"},{"alias_kind":"pith_short_12","alias_value":"RNGRIJL4HVGR","created_at":"2026-05-18T12:31:39.905425+00:00"},{"alias_kind":"pith_short_16","alias_value":"RNGRIJL4HVGRUWOZ","created_at":"2026-05-18T12:31:39.905425+00:00"},{"alias_kind":"pith_short_8","alias_value":"RNGRIJL4","created_at":"2026-05-18T12:31:39.905425+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":32,"internal_anchor_count":12,"sample":[{"citing_arxiv_id":"1907.05984","citing_title":"Distributed Black-Box Optimization via Error Correcting Codes","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"1907.06077","citing_title":"Evolvability ES: Scalable and Direct Optimization of Evolvability","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2410.02543","citing_title":"Diffusion Models are Evolutionary Algorithms","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2502.16156","citing_title":"A Review of Causal Decision Making","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2605.20763","citing_title":"ShapeBench: A Scalable Benchmark and Diagnostic Suite for Standardized Evaluation in Aerodynamic Shape Optimization","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2605.21000","citing_title":"Convergence Analysis of Evolution Strategies for Mixed-Integer Optimization","ref_index":33,"is_internal_anchor":true},{"citing_arxiv_id":"2605.05974","citing_title":"PragLocker: Protecting Agent Intellectual Property in Untrusted Deployments via Non-Portable Prompts","ref_index":127,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18035","citing_title":"New Insight of Variance reduce in Zero-Order Hard-Thresholding: Mitigating Gradient Error and Expansivity Contradictions","ref_index":20,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17394","citing_title":"Stochastic Zeroth-Order Optimization Under Heavy-Tailed Noise","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16844","citing_title":"Artificial Adaptive Intelligence: The Missing Stage Between Narrow and General Intelligence","ref_index":16,"is_internal_anchor":true},{"citing_arxiv_id":"2510.16978","citing_title":"Lark: Biologically Inspired Neuroevolution for Multi-Stakeholder LLM Agents","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2603.06082","citing_title":"Offline Materials Optimization with CliqueFlowmer","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2604.02882","citing_title":"Importance Sampling Optimization with Laplace Principle","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11929","citing_title":"Convergence of zeroth-order proximal point algorithms in the high-temperature regime","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27667","citing_title":"Can Tabular Foundation Models Guide Exploration in Robot Policy Learning?","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21456","citing_title":"Tempered Sequential Monte Carlo for Trajectory and Policy Optimization with Differentiable Dynamics","ref_index":84,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05284","citing_title":"Direct From Darwin: Deriving Advanced Optimizers From Evolutionary First Principles","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09781","citing_title":"Parameter-Efficient Neuroevolution for Diverse LLM Generation: Quality-Diversity Optimization via Prompt Embedding Evolution","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25372","citing_title":"From Cursed to Competitive: Closing the ZO-FO Gap via Input-to-State Stability","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03288","citing_title":"Neural Control: Adjoint Learning Through Equilibrium Constraints","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24708","citing_title":"Scalable Hyperparameter-Divergent Ensemble Training with Automatic Learning Rate Exploration for Large Models","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05974","citing_title":"PragLocker: Protecting Agent Intellectual Property in Untrusted Deployments via Non-Portable Prompts","ref_index":127,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06228","citing_title":"Soft Deterministic Policy Gradient with Gaussian Smoothing","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05284","citing_title":"Direct From Darwin: Deriving Advanced Optimizers From Evolutionary First Principles","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01517","citing_title":"VAnim: Rendering-Aware Sparse State Modeling for Structure-Preserving Vector Animation","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T","json":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T.json","graph_json":"https://pith.science/api/pith-number/RNGRIJL4HVGRUWOZ365I3MB65T/graph.json","events_json":"https://pith.science/api/pith-number/RNGRIJL4HVGRUWOZ365I3MB65T/events.json","paper":"https://pith.science/paper/RNGRIJL4"},"agent_actions":{"view_html":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T","download_json":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T.json","view_paper":"https://pith.science/paper/RNGRIJL4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1703.03864&json=true","fetch_graph":"https://pith.science/api/pith-number/RNGRIJL4HVGRUWOZ365I3MB65T/graph.json","fetch_events":"https://pith.science/api/pith-number/RNGRIJL4HVGRUWOZ365I3MB65T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T/action/storage_attestation","attest_author":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T/action/author_attestation","sign_citation":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T/action/citation_signature","submit_replication":"https://pith.science/pith/RNGRIJL4HVGRUWOZ365I3MB65T/action/replication_record"}},"created_at":"2026-05-18T00:35:46.479492+00:00","updated_at":"2026-05-18T00:35:46.479492+00:00"}