{"id":"https://openalex.org/W7204166068","doi":"https://doi.org/10.48550/arxiv.2608.21898","title":"Training Needs Trustworthy Worlds: Verified Synthetic Web Environments for Agent Learning","display_name":"Training Needs Trustworthy Worlds: Verified Synthetic Web Environments for Agent Learning","publication_year":2026,"publication_date":"2026-08-22","ids":{"openalex":"https://openalex.org/W7204166068","doi":"https://doi.org/10.48550/arxiv.2608.21898"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.21898","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21898","pdf_url":"https://arxiv.org/pdf/2608.21898","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.21898","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5121349031","display_name":"Chenghao Zhang","orcid":"https://orcid.org/0009-0008-0277-2166","observed_orcids":["https://orcid.org/0009-0008-0277-2166"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Chenghao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013618931","display_name":"Yuxi Cheng","orcid":"https://orcid.org/0009-0000-9393-4993","observed_orcids":["https://orcid.org/0009-0000-9393-4993"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cheng, Yuxi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135939873","display_name":"Saisai Hu","orcid":"https://orcid.org/0009-0007-4099-4037","observed_orcids":["https://orcid.org/0009-0007-4099-4037"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Saisai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153416481","display_name":"Canran Xiao","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiao, Canran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027836225","display_name":"Yikai Mao","orcid":"https://orcid.org/0009-0008-4182-3173","observed_orcids":["https://orcid.org/0009-0008-4182-3173"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mao, Yikai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5023802054","display_name":"Dan Roth","orcid":"https://orcid.org/0009-0002-1447-5173","observed_orcids":["https://orcid.org/0009-0002-1447-5173"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Roth, Dan","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/T12016","display_name":"Web Data Mining and Analysis","score":0.7964000105857849,"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":0.7964000105857849,"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":0.14890000224113464,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.0066999997943639755,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/web-agents","display_name":"web agents","score":0.8840000033378601},{"id":"https://openalex.org/keywords/agent-training","display_name":"agent training","score":0.5619999766349792},{"id":"https://openalex.org/keywords/task-feasibility","display_name":"task feasibility","score":0.492000013589859},{"id":"https://openalex.org/keywords/ppo","display_name":"PPO","score":0.4009999930858612},{"id":"https://openalex.org/keywords/webarena","display_name":"WebArena","score":0.5989999771118164}],"concepts":[{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.7193999886512756},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7099000215530396},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.6553999781608582},{"id":"https://openalex.org/C153701036","wikidata":"https://www.wikidata.org/wiki/Q659974","display_name":"Trustworthiness","level":2,"score":0.6284000277519226},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.48420000076293945},{"id":"https://openalex.org/C118643609","wikidata":"https://www.wikidata.org/wiki/Q189210","display_name":"Web application","level":2,"score":0.48019999265670776},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.33660000562667847},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.33309999108314514},{"id":"https://openalex.org/C136764020","wikidata":"https://www.wikidata.org/wiki/Q466","display_name":"World Wide Web","level":1,"score":0.32690000534057617},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32600000500679016},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2957000136375427},{"id":"https://openalex.org/C150899416","wikidata":"https://www.wikidata.org/wiki/Q1820378","display_name":"Transfer of learning","level":2,"score":0.27730000019073486},{"id":"https://openalex.org/C77967617","wikidata":"https://www.wikidata.org/wiki/Q4677561","display_name":"Active learning (machine learning)","level":2,"score":0.2696000039577484}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.21898","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21898","pdf_url":"https://arxiv.org/pdf/2608.21898","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.2608.21898","doi":"https://doi.org/10.48550/arxiv.2608.21898","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.21898","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:2608.21898","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21898","pdf_url":"https://arxiv.org/pdf/2608.21898","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":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204166068.pdf","grobid_xml":"https://content.openalex.org/works/W7204166068.grobid-xml"},"referenced_works_count":11,"referenced_works":["https://openalex.org/W2736601468","https://openalex.org/W2786616516","https://openalex.org/W4392781346","https://openalex.org/W4394776438","https://openalex.org/W4399270457","https://openalex.org/W4402427562","https://openalex.org/W4403580146","https://openalex.org/W4404355279","https://openalex.org/W4415328656","https://openalex.org/W7131050341","https://openalex.org/W7133215859"],"related_works":["https://openalex.org/W7135085063","https://openalex.org/W7129232255","https://openalex.org/W4415346661","https://openalex.org/W7128648738","https://openalex.org/W7204099037","https://openalex.org/W7119478245","https://openalex.org/W7162457869","https://openalex.org/W7220673804","https://openalex.org/W7223027552","https://openalex.org/W4416237351"],"abstract_inverted_index":{"Web":[0],"agents":[1],"promise":[2],"to":[3,133,143],"automate":[4],"complex":[5],"digital":[6],"workflows,":[7],"but":[8],"their":[9],"training":[10,168],"remains":[11],"limited":[12],"by":[13,40],"synthetic":[14,42,117,159,175],"environments":[15,44,118,160],"that":[16,45,157],"look":[17],"plausible":[18],"while":[19,96,135],"hiding":[20],"broken":[21],"links,":[22,67],"inconsistent":[23],"states,":[24],"or":[25],"infeasible":[26],"tasks.":[27],"We":[28],"address":[29],"the":[30],"gap":[31],"between":[32],"scalable":[33,165],"environment":[34],"generation":[35],"and":[36,49,72,77,82,127,140,146,166],"trustworthy":[37],"agent":[38],"learning":[39,177],"constructing":[41],"web":[43,172],"are":[46,93,100],"executable,":[47,182],"auditable,":[48],"grounded":[50],"in":[51],"backend":[52,98],"state.":[53],"Our":[54],"framework":[55],"represents":[56],"each":[57],"generated":[58],"website":[59],"as":[60,163],"a":[61,164],"structured":[62],"scaffold":[63],"of":[64],"pages,":[65],"navigation":[66],"database":[68],"records,":[69],"state-change":[70,105],"markers,":[71,106],"task":[73],"constraints,":[74],"then":[75],"verifies":[76],"repairs":[78],"structural,":[79],"semantic,":[80],"consistency,":[81],"feasibility":[83],"defects":[84,126],"before":[85],"policy":[86],"training.":[87],"During":[88],"interaction,":[89],"ordinary":[90],"UI":[91],"transitions":[92],"executed":[94],"deterministically,":[95],"persistent":[97],"updates":[99],"invoked":[101],"only":[102],"through":[103],"validated":[104],"enabling":[107],"dense":[108],"rewards":[109],"compiled":[110],"from":[111,131,178],"verified":[112,158],"task-progress":[113],"predicates.":[114],"Across":[115],"500":[116],"spanning":[119],"six":[120],"domains,":[121],"our":[122],"method":[123],"reduces":[124],"task-blocking":[125],"improves":[128],"feasible-task":[129],"rate":[130],"48.6%":[132],"94.8%,":[134],"producing":[136],"stronger":[137],"PPO":[138],"policies":[139],"improving":[141],"transfer":[142],"WebArena,":[144],"WebShop,":[145],"MiniWoB++":[147],"without":[148],"LLM":[149],"calls":[150],"at":[151],"evaluation":[152],"time.":[153],"These":[154],"results":[155],"show":[156],"can":[161],"serve":[162],"reliable":[167],"substrate":[169],"for":[170],"compact":[171],"agents,":[173],"shifting":[174],"webagent":[176],"surface-level":[179],"plausibility":[180],"toward":[181],"state-grounded":[183],"supervision.":[184]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-08-26T00:00:00"}