{"id":"https://openalex.org/W7124317355","doi":"https://doi.org/10.65109/opnm7882","title":"Never Worse, Mostly Better: Stable Policy Improvement in Deep Reinforcement Learning","display_name":"Never Worse, Mostly Better: Stable Policy Improvement in Deep Reinforcement Learning","publication_year":2023,"publication_date":"2023-05-30","ids":{"openalex":"https://openalex.org/W7124317355","doi":"https://doi.org/10.65109/opnm7882"},"language":"en","primary_location":{"id":"doi:10.65109/opnm7882","doi":"https://doi.org/10.65109/opnm7882","is_oa":false,"landing_page_url":"https://doi.org/10.65109/opnm7882","pdf_url":null,"source":{"id":"https://openalex.org/S4306417597","display_name":"Adaptive Agents and Multi-Agents Systems","issn_l":"1548-8403","issn":["1548-8403"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"International Joint Conference on Autonomous Agents and Multiagent Systems","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","is_traditional":true,"indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/1910.01062","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5046403197","display_name":"Pranav Khanna","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I64295750","display_name":"Indian Institute of Technology Indore","ror":"https://ror.org/01hhf7w52","country_code":"IN","type":"education","lineage":["https://openalex.org/I64295750"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Pranav Khanna","raw_affiliation_strings":["Indian Institute of Technology, Indore, India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Indian Institute of Technology, Indore, India","institution_ids":["https://openalex.org/I64295750"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049376842","display_name":"Guy Tennenholtz","orcid":"https://orcid.org/0000-0002-2727-2351","observed_orcids":["https://orcid.org/0000-0002-2727-2351"]},"institutions":[{"id":"https://openalex.org/I174306211","display_name":"Technion \u2013 Israel Institute of Technology","ror":"https://ror.org/03qryx823","country_code":"IL","type":"education","lineage":["https://openalex.org/I174306211"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Guy Tennenholtz","raw_affiliation_strings":["Technion, Haifa, Israel"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Technion, Haifa, Israel","institution_ids":["https://openalex.org/I174306211"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5018784842","display_name":"Nadav Merlis","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I174306211","display_name":"Technion \u2013 Israel Institute of Technology","ror":"https://ror.org/03qryx823","country_code":"IL","type":"education","lineage":["https://openalex.org/I174306211"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Nadav Merlis","raw_affiliation_strings":["Technion, Haifa, Israel"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Technion, Haifa, Israel","institution_ids":["https://openalex.org/I174306211"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036260775","display_name":"Shie Mannor","orcid":"https://orcid.org/0000-0003-4439-7647","observed_orcids":["https://orcid.org/0000-0003-4439-7647"]},"institutions":[{"id":"https://openalex.org/I174306211","display_name":"Technion \u2013 Israel Institute of Technology","ror":"https://ror.org/03qryx823","country_code":"IL","type":"education","lineage":["https://openalex.org/I174306211"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Shie Mannor","raw_affiliation_strings":["Technion & NVIDIA, Haifa, Israel"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Technion & NVIDIA, Haifa, Israel","institution_ids":["https://openalex.org/I174306211"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5002549620","display_name":"Chen Tessler","orcid":"https://orcid.org/0000-0001-6447-9864","observed_orcids":["https://orcid.org/0000-0001-6447-9864"]},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chen Tessler","raw_affiliation_strings":["NVIDIA, Haifa, Israel"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA, Haifa, Israel","institution_ids":["https://openalex.org/I4210127875"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.1105,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":{"value":0.26502952,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":88,"max":93},"biblio":{"volume":null,"issue":null,"first_page":"2430","last_page":"2432"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12794","display_name":"Adaptive Dynamic Programming Control","score":0.0,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11574","display_name":"Artificial Intelligence in Games","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/deep-reinforcement-learning","display_name":"deep reinforcement learning","score":0.9049999713897705},{"id":"https://openalex.org/keywords/confidence-limits","display_name":"confidence limits","score":0.7549999952316284},{"id":"https://openalex.org/keywords/mount-everest","display_name":"Mount Everest","score":0.8109999895095825},{"id":"https://openalex.org/keywords/continuous-control","display_name":"continuous control","score":0.7639999985694885}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7998999953269958},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.744700014591217},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6607000231742859},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.5860999822616577},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5467000007629395},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5248000025749207},{"id":"https://openalex.org/C136197465","wikidata":"https://www.wikidata.org/wiki/Q1729295","display_name":"Variety (cybernetics)","level":2,"score":0.4756999909877777},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.4487999975681305},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.358599990606308},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.3416000008583069},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.31779998540878296},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3158000111579895},{"id":"https://openalex.org/C196340769","wikidata":"https://www.wikidata.org/wiki/Q7698910","display_name":"Temporal difference learning","level":3,"score":0.2799000144004822},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.2678000032901764},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.2660999894142151},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.25270000100135803}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.65109/opnm7882","doi":"https://doi.org/10.65109/opnm7882","is_oa":false,"landing_page_url":"https://doi.org/10.65109/opnm7882","pdf_url":null,"source":{"id":"https://openalex.org/S4306417597","display_name":"Adaptive Agents and Multi-Agents Systems","issn_l":"1548-8403","issn":["1548-8403"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"International Joint Conference on Autonomous Agents and Multiagent Systems","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:1910.01062","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/1910.01062","pdf_url":"https://arxiv.org/pdf/1910.01062","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.1910.01062","doi":"https://doi.org/10.48550/arxiv.1910.01062","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.1910.01062","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:1910.01062","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/1910.01062","pdf_url":"https://arxiv.org/pdf/1910.01062","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7124317355.pdf","grobid_xml":"https://content.openalex.org/works/W7124317355.grobid-xml"},"referenced_works_count":7,"referenced_works":["https://openalex.org/W2616964725","https://openalex.org/W2736601468","https://openalex.org/W2768908787","https://openalex.org/W2784465508","https://openalex.org/W2787938642","https://openalex.org/W3093206925","https://openalex.org/W4225495521"],"related_works":["https://openalex.org/W3004986541","https://openalex.org/W2980250066","https://openalex.org/W2979064149","https://openalex.org/W7133240016","https://openalex.org/W4280647283","https://openalex.org/W2596758708","https://openalex.org/W2964108292","https://openalex.org/W3144527084","https://openalex.org/W7128616416","https://openalex.org/W2892076218"],"abstract_inverted_index":{"In":[0,72],"recent":[1],"years,":[2],"there":[3],"has":[4,29],"been":[5,30],"significant":[6,134],"progress":[7],"in":[8,46,126,136],"applying":[9],"deep":[10],"reinforcement":[11],"learning":[12,82],"(RL)":[13],"for":[14,78],"solving":[15],"challenging":[16],"problems":[17],"across":[18],"a":[19,96,112],"wide":[20],"variety":[21],"of":[22,26,70,111,123,128],"domains.":[23],"Nevertheless,":[24],"convergence":[25],"various":[27],"methods":[28,77],"shown":[31],"to":[32,37,90],"suffer":[33],"from":[34],"inconsistencies,":[35],"due":[36],"algorithmic":[38],"instability":[39],"and":[40,131,139],"variance,":[41],"as":[42,44],"well":[43],"stochasticity":[45],"the":[47,52,55,80,92,121],"benchmark":[48,98],"environments.":[49],"Particularly,":[50],"despite":[51],"fact":[53],"that":[54],"agent's":[56,81],"performance":[57,130],"may":[58,64],"be":[59],"improving":[60],"on":[61],"average,":[62],"it":[63],"abruptly":[65],"deteriorate":[66],"at":[67],"late":[68],"stages":[69],"training.":[71],"this":[73],"work,":[74],"we":[75,119],"study":[76],"enhancing":[79],"process,":[83],"by":[84],"providing":[85],"conservative":[86],"updates":[87],"with":[88,133],"respect":[89],"either":[91],"obtained":[93],"history":[94],"or":[95],"reference":[97,113],"policy.":[99,114],"Our":[100],"method,":[101],"termed":[102],"EVEREST,":[103],"obtains":[104],"high":[105],"confidence":[106,109],"improvements":[107,135],"via":[108],"bounds":[110],"Through":[115],"extensive":[116],"empirical":[117],"analysis":[118],"demonstrate":[120],"benefit":[122],"our":[124],"approach":[125],"terms":[127],"both":[129],"stabilization,":[132],"continuous":[137],"control":[138],"Atari":[140],"benchmarks.":[141]},"counts_by_year":[{"year":2023,"cited_by_count":1}],"updated_date":"2026-11-10T11:39:43.021295","created_date":"2026-01-16T00:00:00"}