{"id":"https://openalex.org/W2896930824","doi":"https://doi.org/10.48550/arxiv.1810.08575","title":"Supervising strong learners by amplifying weak experts","display_name":"Supervising strong learners by amplifying weak experts","publication_year":2018,"publication_date":"2018-10-19","ids":{"openalex":"https://openalex.org/W2896930824","doi":"https://doi.org/10.48550/arxiv.1810.08575","mag":"2896930824"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:1810.08575","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/1810.08575","pdf_url":"https://arxiv.org/pdf/1810.08575","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"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/1810.08575","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5037736834","display_name":"Paul F. Christiano","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Christiano, Paul","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050408969","display_name":"Buck Shlegeris","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shlegeris, Buck","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5066197394","display_name":"Dario Amodei","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Amodei, Dario","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":39,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9843999743461609,"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":0.9843999743461609,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.00839999970048666,"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/T11574","display_name":"Artificial Intelligence in Games","score":0.002099999925121665,"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/human-in-the-loop-machine-learning","display_name":"human-in-the-loop learning","score":0.5740000009536743},{"id":"https://openalex.org/keywords/subproblem-decomposition","display_name":"subproblem decomposition","score":0.675000011920929},{"id":"https://openalex.org/keywords/algorithmic-environments","display_name":"algorithmic environments","score":0.5870000123977661}],"concepts":[{"id":"https://openalex.org/C140479938","wikidata":"https://www.wikidata.org/wiki/Q5254619","display_name":"Iterated function","level":2,"score":0.87149977684021},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7179123759269714},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.7044578194618225},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.48616668581962585},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.44579389691352844},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4289611577987671},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4211328327655792},{"id":"https://openalex.org/C2780148112","wikidata":"https://www.wikidata.org/wiki/Q1432581","display_name":"Proxy (statistics)","level":2,"score":0.4141013026237488},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.15250778198242188},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.11723434925079346},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C78458016","wikidata":"https://www.wikidata.org/wiki/Q840400","display_name":"Evolutionary biology","level":1,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"pmh:oai:arXiv.org:1810.08575","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/1810.08575","pdf_url":"https://arxiv.org/pdf/1810.08575","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":"mag:2896930824","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/pdf/1810.08575.pdf","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},{"id":"doi:10.48550/arxiv.1810.08575","doi":"https://doi.org/10.48550/arxiv.1810.08575","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.1810.08575","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:1810.08575","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/1810.08575","pdf_url":"https://arxiv.org/pdf/1810.08575","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":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":18,"referenced_works":["https://openalex.org/W1999874108","https://openalex.org/W2061562262","https://openalex.org/W2098584016","https://openalex.org/W2156869222","https://openalex.org/W2173051530","https://openalex.org/W2214429195","https://openalex.org/W2290104316","https://openalex.org/W2462906003","https://openalex.org/W2578937396","https://openalex.org/W2607964821","https://openalex.org/W2618097077","https://openalex.org/W2626778328","https://openalex.org/W2626804490","https://openalex.org/W2766447205","https://openalex.org/W2772709170","https://openalex.org/W2792012198","https://openalex.org/W2798877128","https://openalex.org/W2951952563"],"related_works":["https://openalex.org/W7095429339","https://openalex.org/W2604325524","https://openalex.org/W2142641780","https://openalex.org/W3034343528","https://openalex.org/W3012995883","https://openalex.org/W4382318875","https://openalex.org/W7095761615","https://openalex.org/W2462704484","https://openalex.org/W2991758148","https://openalex.org/W7096054825"],"abstract_inverted_index":{"Many":[0],"real":[1],"world":[2],"learning":[3],"tasks":[4],"involve":[5],"complex":[6,115],"or":[7,20,35],"hard-to-specify":[8],"objectives,":[9],"and":[10],"using":[11],"an":[12,58],"easier-to-specify":[13],"proxy":[14],"can":[15,112],"lead":[16],"to":[17,26,51,75,83],"poor":[18],"performance":[19],"misaligned":[21],"behavior.":[22],"One":[23],"solution":[24],"is":[25,45,80],"have":[27],"humans":[28],"provide":[29],"a":[30,49,66],"training":[31,60,67],"signal":[32,68],"by":[33,72],"demonstrating":[34],"judging":[36],"performance,":[37],"but":[38],"this":[39],"approach":[40],"fails":[41],"if":[42],"the":[43],"task":[44],"too":[46],"complicated":[47],"for":[48,69],"human":[50],"directly":[52],"evaluate.":[53],"We":[54,102],"propose":[55],"Iterated":[56,78,110],"Amplification,":[57],"alternative":[59],"strategy":[61],"which":[62],"progressively":[63],"builds":[64],"up":[65],"difficult":[70],"problems":[71],"combining":[73],"solutions":[74],"easier":[76],"subproblems.":[77],"Amplification":[79,111],"closely":[81],"related":[82],"Expert":[84],"Iteration":[85],"(Anthony":[86],"et":[87,91],"al.,":[88,92],"2017;":[89],"Silver":[90],"2017),":[93],"except":[94],"that":[95,109],"it":[96],"uses":[97],"no":[98],"external":[99],"reward":[100],"function.":[101],"present":[103],"results":[104],"in":[105],"algorithmic":[106],"environments,":[107],"showing":[108],"efficiently":[113],"learn":[114],"behaviors.":[116]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":5},{"year":2022,"cited_by_count":2},{"year":2021,"cited_by_count":7},{"year":2020,"cited_by_count":5},{"year":2019,"cited_by_count":9},{"year":2018,"cited_by_count":2}],"updated_date":"2026-11-03T10:50:19.383937","created_date":"2018-10-26T00:00:00"}