{"id":"https://openalex.org/W4388717834","doi":"https://doi.org/10.48550/arxiv.2311.07723","title":"Generalization Analogies: A Testbed for Generalizing AI Oversight to Hard-To-Measure Domains","display_name":"Generalization Analogies: A Testbed for Generalizing AI Oversight to Hard-To-Measure Domains","publication_year":2023,"publication_date":"2023-11-13","ids":{"openalex":"https://openalex.org/W4388717834","doi":"https://doi.org/10.48550/arxiv.2311.07723"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2311.07723","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2311.07723","pdf_url":"https://arxiv.org/pdf/2311.07723","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","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/2311.07723","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5072000814","display_name":"Joshua Clymer","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Clymer, Joshua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Baker, Garrett","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Baker, Garrett","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104198742","display_name":"Rohan Subramani","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Subramani, Rohan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5040090483","display_name":"Sam C. Wang","orcid":"https://orcid.org/0000-0003-3626-5273","observed_orcids":["https://orcid.org/0000-0003-3626-5273"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Sam","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":2,"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/T10883","display_name":"Ethics and Social Impacts of AI","score":0.711899995803833,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T10883","display_name":"Ethics and Social Impacts of AI","score":0.711899995803833,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.042399998754262924,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.039900001138448715,"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/distribution-shift","display_name":"distribution shift","score":0.6349999904632568},{"id":"https://openalex.org/keywords/human-oversight-of-ai","display_name":"human oversight of AI","score":0.6449999809265137},{"id":"https://openalex.org/keywords/human-feedback","display_name":"human feedback","score":0.45500001311302185},{"id":"https://openalex.org/keywords/instruction-following","display_name":"instruction following","score":0.6439999938011169},{"id":"https://openalex.org/keywords/personas","display_name":"personas","score":0.48500001430511475},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8550000190734863}],"concepts":[{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.8777629137039185},{"id":"https://openalex.org/C130440534","wikidata":"https://www.wikidata.org/wiki/Q14946528","display_name":"Conflation","level":2,"score":0.756121814250946},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6911786794662476},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.624254584312439},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5289036631584167},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.49745967984199524},{"id":"https://openalex.org/C31395832","wikidata":"https://www.wikidata.org/wiki/Q1318674","display_name":"Testbed","level":2,"score":0.4786730706691742},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41453635692596436},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.3466480076313019},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.2999415397644043},{"id":"https://openalex.org/C111472728","wikidata":"https://www.wikidata.org/wiki/Q9471","display_name":"Epistemology","level":1,"score":0.16378724575042725},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.12739869952201843},{"id":"https://openalex.org/C136764020","wikidata":"https://www.wikidata.org/wiki/Q466","display_name":"World Wide Web","level":1,"score":0.09146678447723389},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2311.07723","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2311.07723","pdf_url":"https://arxiv.org/pdf/2311.07723","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2311.07723","doi":"https://doi.org/10.48550/arxiv.2311.07723","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2311.07723","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2311.07723","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2311.07723","pdf_url":"https://arxiv.org/pdf/2311.07723","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","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/W4388717834.pdf","grobid_xml":"https://content.openalex.org/works/W4388717834.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4389115197","https://openalex.org/W7169671228","https://openalex.org/W4415797097","https://openalex.org/W4416073615","https://openalex.org/W4416101722","https://openalex.org/W4399795554","https://openalex.org/W4415109460","https://openalex.org/W7160943590","https://openalex.org/W4416288891","https://openalex.org/W7222913869"],"abstract_inverted_index":{"As":[0],"AI":[1],"systems":[2],"become":[3],"more":[4,10],"intelligent":[5],"and":[6,78],"their":[7],"behavior":[8],"becomes":[9],"challenging":[11,114],"to":[12,17,28,44,73,103],"assess,":[13],"they":[14],"may":[15],"learn":[16,72],"game":[18],"the":[19,111,118],"flaws":[20],"of":[21,25],"human":[22,42],"feedback":[23,43],"instead":[24,79],"genuinely":[26],"striving":[27],"follow":[29],"instructions;":[30],"however,":[31],"this":[32],"risk":[33],"can":[34],"be":[35],"mitigated":[36],"by":[37,76],"controlling":[38,130],"how":[39,53],"LLMs":[40],"generalize":[41],"situations":[45],"where":[46],"it":[47],"is":[48],"unreliable.":[49],"To":[50],"better":[51,94],"understand":[52],"reward":[54,68,89,131],"models":[55,69],"generalize,":[56],"we":[57,124],"craft":[58],"69":[59],"distribution":[60,115],"shifts":[61,116],"spanning":[62],"8":[63],"categories.":[64],"We":[65,109],"find":[66],"that":[67,82],"do":[70],"not":[71],"evaluate":[74],"`instruction-following'":[75],"default":[77],"favor":[80],"personas":[81],"resemble":[83],"internet":[84],"text.":[85],"Techniques":[86],"for":[87],"interpreting":[88],"models'":[90],"internal":[91],"representations":[92],"achieve":[93],"generalization":[95],"than":[96],"standard":[97],"fine-tuning,":[98],"but":[99],"still":[100],"frequently":[101],"fail":[102],"distinguish":[104],"instruction-following":[105],"from":[106],"conflated":[107],"behaviors.":[108],"consolidate":[110],"15":[112],"most":[113],"into":[117],"GENeralization":[119],"analogIES":[120],"(GENIES)":[121],"benchmark,":[122],"which":[123],"hope":[125],"will":[126],"enable":[127],"progress":[128],"toward":[129],"model":[132],"generalization.":[133]},"counts_by_year":[{"year":2024,"cited_by_count":2}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2023-11-16T00:00:00"}