{"id":"https://openalex.org/W4415345646","doi":"https://doi.org/10.52202/085713-1774","title":"How to Train Your LLM Web Agent: A Statistical Diagnosis","display_name":"How to Train Your LLM Web Agent: A Statistical Diagnosis","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4415345646","doi":"https://doi.org/10.52202/085713-1774"},"language":"en","primary_location":{"id":"doi:10.52202/085713-1774","doi":"https://doi.org/10.52202/085713-1774","is_oa":false,"landing_page_url":"https://doi.org/10.52202/085713-1774","pdf_url":null,"source":{"id":"https://openalex.org/S4363606243","display_name":"neural information processing systems","issn_l":"1049-5258","issn":["1049-5258"],"is_oa":false,"is_in_doaj":false,"is_core":true,"listed_in":["cwts-core","norway-2"],"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":"Advances in Neural Information Processing Systems 38","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/2507.04103","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5092905548","display_name":"Dheeraj Vattikonda","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210164802","display_name":"Mila - Quebec Artificial Intelligence Institute","ror":"https://ror.org/05c22rx21","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210164802"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Dheeraj Vattikonda","raw_affiliation_strings":["Mila"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mila","institution_ids":["https://openalex.org/I4210164802"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Santhoshi Ravichandran","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210164802","display_name":"Mila - Quebec Artificial Intelligence Institute","ror":"https://ror.org/05c22rx21","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210164802"]},{"id":"https://openalex.org/I70931966","display_name":"Universit\u00e9 de Montr\u00e9al","ror":"https://ror.org/0161xgx34","country_code":"CA","type":"education","lineage":["https://openalex.org/I70931966"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Santhoshi Ravichandran","raw_affiliation_strings":["University of Montreal (MILA)"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Montreal (MILA)","institution_ids":["https://openalex.org/I4210164802","https://openalex.org/I70931966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023766910","display_name":"Emiliano Penaloza","orcid":"https://orcid.org/0000-0003-3785-6796","observed_orcids":["https://orcid.org/0000-0003-3785-6796"]},"institutions":[{"id":"https://openalex.org/I4210164802","display_name":"Mila - Quebec Artificial Intelligence Institute","ror":"https://ror.org/05c22rx21","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210164802"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Emiliano Penaloza","raw_affiliation_strings":["Mila - Qu\u00e9bec, ServiceNow"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mila - Qu\u00e9bec, ServiceNow","institution_ids":["https://openalex.org/I4210164802"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5059233651","display_name":"Hadi Nekoei","orcid":"https://orcid.org/0000-0002-9755-6659","observed_orcids":["https://orcid.org/0000-0002-9755-6659"]},"institutions":[{"id":"https://openalex.org/I4210164802","display_name":"Mila - Quebec Artificial Intelligence Institute","ror":"https://ror.org/05c22rx21","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210164802"]},{"id":"https://openalex.org/I70931966","display_name":"Universit\u00e9 de Montr\u00e9al","ror":"https://ror.org/0161xgx34","country_code":"CA","type":"education","lineage":["https://openalex.org/I70931966"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Hadi Nekoei","raw_affiliation_strings":["Mila, University of Montreal"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mila, University of Montreal","institution_ids":["https://openalex.org/I4210164802","https://openalex.org/I70931966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100493934","display_name":"Thibault Le Sellier De Chezelles","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I45683168","display_name":"Polytechnique Montr\u00e9al","ror":"https://ror.org/05f8d4e86","country_code":"CA","type":"education","lineage":["https://openalex.org/I45683168"]},{"id":"https://openalex.org/I70931966","display_name":"Universit\u00e9 de Montr\u00e9al","ror":"https://ror.org/0161xgx34","country_code":"CA","type":"education","lineage":["https://openalex.org/I70931966"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Thibault de Chezelles","raw_affiliation_strings":["\u00c9cole Polytechnique de Montr\u00e9al, Universit\u00e9 de Montr\u00e9al"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"\u00c9cole Polytechnique de Montr\u00e9al, Universit\u00e9 de Montr\u00e9al","institution_ids":["https://openalex.org/I45683168","https://openalex.org/I70931966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045032553","display_name":"Megh Thakkar","orcid":"https://orcid.org/0009-0004-8224-622X","observed_orcids":["https://orcid.org/0009-0004-8224-622X"]},"institutions":[{"id":"https://openalex.org/I70931966","display_name":"Universit\u00e9 de Montr\u00e9al","ror":"https://ror.org/0161xgx34","country_code":"CA","type":"education","lineage":["https://openalex.org/I70931966"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Megh Thakkar","raw_affiliation_strings":["Universit\u00e9 de Montr\u00e9al"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Universit\u00e9 de Montr\u00e9al","institution_ids":["https://openalex.org/I70931966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5025987694","display_name":"Nicolas Gontier","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210164802","display_name":"Mila - Quebec Artificial Intelligence Institute","ror":"https://ror.org/05c22rx21","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210164802"]},{"id":"https://openalex.org/I45683168","display_name":"Polytechnique Montr\u00e9al","ror":"https://ror.org/05f8d4e86","country_code":"CA","type":"education","lineage":["https://openalex.org/I45683168"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Nicolas Gontier","raw_affiliation_strings":["Mila, Polytechnique Montr\u00e9al"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mila, Polytechnique Montr\u00e9al","institution_ids":["https://openalex.org/I4210164802","https://openalex.org/I45683168"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005112777","display_name":"Miguel Mu\u00f1oz-M\u00e1rmol","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210167039","display_name":"ServiceNow (United States)","ror":"https://ror.org/05xr0bc04","country_code":"US","type":"company","lineage":["https://openalex.org/I4210167039"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Miguel Mu\u00f1oz-M\u00e1rmol","raw_affiliation_strings":["ServiceNow Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ServiceNow Inc","institution_ids":["https://openalex.org/I4210167039"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5115086899","display_name":"Sahar Omidi Shayegan","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sahar Omidi Shayegan","raw_affiliation_strings":["Veris AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Veris AI","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102520041","display_name":"Stefania Raimondo","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210167039","display_name":"ServiceNow (United States)","ror":"https://ror.org/05xr0bc04","country_code":"US","type":"company","lineage":["https://openalex.org/I4210167039"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Stefania Raimondo","raw_affiliation_strings":["Service Now Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Service Now Research","institution_ids":["https://openalex.org/I4210167039"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100372152","display_name":"Xue Liu","orcid":"https://orcid.org/0000-0001-5252-3442","observed_orcids":["https://orcid.org/0000-0001-5252-3442"]},"institutions":[{"id":"https://openalex.org/I4210113480","display_name":"Mohamed bin Zayed University of Artificial Intelligence","ror":"https://ror.org/0258gkt32","country_code":"AE","type":"education","lineage":["https://openalex.org/I4210113480"]}],"countries":["AE"],"is_corresponding":false,"raw_author_name":"Steve (Xue) Liu","raw_affiliation_strings":["MBZUAI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"MBZUAI","institution_ids":["https://openalex.org/I4210113480"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5028239106","display_name":"Alexandre Drouin","orcid":"https://orcid.org/0000-0001-7718-0319","observed_orcids":["https://orcid.org/0000-0001-7718-0319"]},"institutions":[{"id":"https://openalex.org/I4210164802","display_name":"Mila - Quebec Artificial Intelligence Institute","ror":"https://ror.org/05c22rx21","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210164802"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Alexandre Drouin","raw_affiliation_strings":["ServiceNow Research / Mila"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ServiceNow Research / Mila","institution_ids":["https://openalex.org/I4210164802"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107175005","display_name":"Alexandre Pich\u00e9","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210164802","display_name":"Mila - Quebec Artificial Intelligence Institute","ror":"https://ror.org/05c22rx21","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210164802"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Alexandre Piche","raw_affiliation_strings":["Mila"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mila","institution_ids":["https://openalex.org/I4210164802"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107175005","display_name":"Alexandre Pich\u00e9","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210167039","display_name":"ServiceNow (United States)","ror":"https://ror.org/05xr0bc04","country_code":"US","type":"company","lineage":["https://openalex.org/I4210167039"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Alexandre Lacoste","raw_affiliation_strings":["Service Now Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Service Now Research","institution_ids":["https://openalex.org/I4210167039"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5130023931","display_name":"Massimo Caccia","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210167039","display_name":"ServiceNow (United States)","ror":"https://ror.org/05xr0bc04","country_code":"US","type":"company","lineage":["https://openalex.org/I4210167039"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Massimo Caccia","raw_affiliation_strings":["ServiceNow Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ServiceNow Research","institution_ids":["https://openalex.org/I4210167039"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":8.7725,"has_fulltext":true,"cited_by_count":16,"citation_normalized_percentile":{"value":0.97991678,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"59259","last_page":"59285"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12016","display_name":"Web Data Mining and Analysis","score":0.6952999830245972,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T12016","display_name":"Web Data Mining and Analysis","score":0.6952999830245972,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.1363999992609024,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.01600000075995922,"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/supervised-fine-tuning","display_name":"supervised fine-tuning","score":0.9129999876022339},{"id":"https://openalex.org/keywords/on-policy-learning","display_name":"on-policy learning","score":0.9950000047683716},{"id":"https://openalex.org/keywords/hyperparameter-sensitivity","display_name":"hyperparameter sensitivity","score":0.6629999876022339}],"concepts":[{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.6894000172615051},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6669999957084656},{"id":"https://openalex.org/C207609745","wikidata":"https://www.wikidata.org/wiki/Q4944086","display_name":"Bootstrapping (finance)","level":2,"score":0.5986999869346619},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5475999712944031},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5264000296592712},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.5189999938011169},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5026999711990356},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.490200012922287},{"id":"https://openalex.org/C194648553","wikidata":"https://www.wikidata.org/wiki/Q1364774","display_name":"Spare part","level":2,"score":0.45739999413490295},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.4507000148296356},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.42640000581741333},{"id":"https://openalex.org/C2779955035","wikidata":"https://www.wikidata.org/wiki/Q4686785","display_name":"Advice (programming)","level":2,"score":0.3465000092983246},{"id":"https://openalex.org/C118643609","wikidata":"https://www.wikidata.org/wiki/Q189210","display_name":"Web application","level":2,"score":0.3377000093460083},{"id":"https://openalex.org/C137635306","wikidata":"https://www.wikidata.org/wiki/Q182667","display_name":"Pareto principle","level":2,"score":0.3292999863624573},{"id":"https://openalex.org/C110875604","wikidata":"https://www.wikidata.org/wiki/Q75","display_name":"The Internet","level":2,"score":0.31189998984336853},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.27639999985694885},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.27079999446868896},{"id":"https://openalex.org/C129848803","wikidata":"https://www.wikidata.org/wiki/Q2564360","display_name":"Sample size determination","level":2,"score":0.2689000070095062},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.26420000195503235},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.26330000162124634},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.26030001044273376},{"id":"https://openalex.org/C87007009","wikidata":"https://www.wikidata.org/wiki/Q210832","display_name":"Statistical hypothesis testing","level":2,"score":0.25600001215934753},{"id":"https://openalex.org/C198352243","wikidata":"https://www.wikidata.org/wiki/Q37105","display_name":"Line (geometry)","level":2,"score":0.2558000087738037}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.52202/085713-1774","doi":"https://doi.org/10.52202/085713-1774","is_oa":false,"landing_page_url":"https://doi.org/10.52202/085713-1774","pdf_url":null,"source":{"id":"https://openalex.org/S4363606243","display_name":"neural information processing systems","issn_l":"1049-5258","issn":["1049-5258"],"is_oa":false,"is_in_doaj":false,"is_core":true,"listed_in":["cwts-core","norway-2"],"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":"Advances in Neural Information Processing Systems 38","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2507.04103","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2507.04103","pdf_url":"https://arxiv.org/pdf/2507.04103","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":"pmh:oai:publications.polymtl.ca:80333","doi":null,"is_oa":false,"landing_page_url":"https://publications.polymtl.ca/80333/","pdf_url":null,"source":{"id":"https://openalex.org/S4306401013","display_name":"PolyPublie (\u00c9cole Polytechnique de Montr\u00e9al)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I45683168","host_organization_name":"Polytechnique Montr\u00e9al","host_organization_lineage":["https://openalex.org/I45683168"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Vattikonda, D., Ravichandran, S., Penaloza, E., Nekoei, H., Le Sellier de Chezelles, T., Thakkar, M., Angelard-Gontier, N., Mu\u00f1oz-M\u00e1rmol, M., Omidi Shayegan, S., Raimondo, S., Liu, X., Drouin, A., Pich\u00e9, A., & Caccia, M. (2025, December). How to Train Your LLM Web Agent: A Statistical Diagnosis [Paper]. 39th Conference on Neural Information Processing Systems (NeurIPS 2025), San Diego, CA, USA.","raw_type":"http://purl.org/coar/resource_type/c_5794"},{"id":"doi:10.48550/arxiv.2507.04103","doi":"https://doi.org/10.48550/arxiv.2507.04103","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2507.04103","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:2507.04103","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2507.04103","pdf_url":"https://arxiv.org/pdf/2507.04103","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/W4415345646.pdf","grobid_xml":"https://content.openalex.org/works/W4415345646.grobid-xml"},"referenced_works_count":17,"referenced_works":["https://openalex.org/W1995945562","https://openalex.org/W2786616516","https://openalex.org/W3157752112","https://openalex.org/W4233413206","https://openalex.org/W4385327683","https://openalex.org/W4403885431","https://openalex.org/W4405627213","https://openalex.org/W4406961529","https://openalex.org/W4409049124","https://openalex.org/W4414900865","https://openalex.org/W4414903627","https://openalex.org/W4414973451","https://openalex.org/W4415134985","https://openalex.org/W4415288811","https://openalex.org/W4415798926","https://openalex.org/W4417522340","https://openalex.org/W7196953391"],"related_works":["https://openalex.org/W4404355279","https://openalex.org/W4403580146","https://openalex.org/W4405263848","https://openalex.org/W4417314204","https://openalex.org/W7101589181","https://openalex.org/W7140090787","https://openalex.org/W7152963405","https://openalex.org/W6891936057","https://openalex.org/W4416451824","https://openalex.org/W7214162073"],"abstract_inverted_index":{"LLM-based":[0,56],"web":[1,45,57],"agents":[2],"have":[3],"recently":[4],"made":[5],"significant":[6],"progress,":[7],"but":[8],"much":[9],"of":[10,43,159,167],"it":[11],"has":[12,24],"occurred":[13],"in":[14],"closed-source":[15,189],"systems,":[16],"widening":[17],"the":[18,41,49,64,160,164,174,180,186],"gap":[19,187],"with":[20,140,188],"open-source":[21],"alternatives.":[22],"Progress":[23],"been":[25],"held":[26],"back":[27],"by":[28,100],"two":[29],"key":[30],"challenges:":[31],"first,":[32],"a":[33,79,83,90],"narrow":[34],"focus":[35],"on":[36,69,148,170],"single-step":[37],"tasks":[38],"that":[39,137,183],"overlooks":[40],"complexity":[42],"multi-step":[44],"interactions;":[46],"and":[47,127,151,178],"second,":[48],"high":[50],"compute":[51,70,161],"costs":[52],"required":[53],"to":[54,88,110,130,162],"post-train":[55],"agents.":[58],"To":[59,117],"address":[60],"this,":[61],"we":[62,123],"present":[63],"first":[65],"statistically":[66],"grounded":[67],"study":[68],"allocation":[71],"for":[72],"LLM":[73],"web-agent":[74],"post-training.":[75],"Our":[76,134],"approach":[77,146],"uses":[78],"two-stage":[80],"pipeline,":[81],"training":[82],"Llama":[84,91],"3.1":[85],"8B":[86],"student":[87],"imitate":[89],"3.3":[92],"70B":[93],"teacher":[94],"via":[95],"supervised":[96],"fine-tuning":[97],"(SFT),":[98],"followed":[99],"on-policy":[101,141],"reinforcement":[102],"learning.":[103],"We":[104],"find":[105],"this":[106,154],"process":[107],"highly":[108],"sensitive":[109],"hyperparameter":[111],"choices,":[112],"making":[113],"exhaustive":[114],"sweeps":[115],"impractical.":[116],"spare":[118],"others":[119],"from":[120],"expensive":[121],"trial-and-error,":[122],"sample":[124],"1,370":[125],"configurations":[126],"use":[128],"bootstrapping":[129],"estimate":[131],"effective":[132],"hyperparameters.":[133],"results":[135],"show":[136],"combining":[138],"SFT":[139,169],"RL":[142],"consistently":[143],"outperforms":[144],"either":[145],"alone":[147],"both":[149],"WorkArena":[150],"MiniWob++.":[152],"Further,":[153],"strategy":[155,182],"requires":[156],"only":[157,181],"55%":[158],"match":[163],"peak":[165],"performance":[166],"pure":[168],"MiniWob++,":[171],"effectively":[172],"pushing":[173],"compute-performance":[175],"Pareto":[176],"frontier,":[177],"is":[179],"can":[184],"close":[185],"models.":[190]},"counts_by_year":[{"year":2026,"cited_by_count":12},{"year":2025,"cited_by_count":4}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-20T00:00:00"}