{"id":"https://openalex.org/W7165380398","doi":"https://doi.org/10.48550/arxiv.2606.19930","title":"MobileForge: Annotation-Free Adaptation for Mobile GUI Agents with Hierarchical Feedback-Guided Policy Optimization","display_name":"MobileForge: Annotation-Free Adaptation for Mobile GUI Agents with Hierarchical Feedback-Guided Policy Optimization","publication_year":2026,"publication_date":"2026-06-18","ids":{"openalex":"https://openalex.org/W7165380398","doi":"https://doi.org/10.48550/arxiv.2606.19930"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.19930","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19930","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.19930","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5138994028","display_name":"Guangyi Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Guangyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010614508","display_name":"Pengxiang Zhao","orcid":"https://orcid.org/0000-0002-5279-9331"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Pengxiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125391623","display_name":"Gao Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Gao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112398353","display_name":"Yiwen Yin","orcid":"https://orcid.org/0009-0002-0620-7902"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yin, Yiwen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5060424860","display_name":"Mading Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Mading","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139018980","display_name":"Liang Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Liang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019036278","display_name":"Congxiao Liu","orcid":"https://orcid.org/0000-0001-8871-4195"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Congxiao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064853949","display_name":"\u5f20\u7426 Zhang Qi","orcid":"https://orcid.org/0000-0003-0058-9022"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qi, Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139008674","display_name":"Mengyan Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Mengyan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139020072","display_name":"Liang Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Liang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139009614","display_name":"Yong Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Yong","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":false,"cited_by_count":0,"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/T10639","display_name":"Advanced Software Engineering Methodologies","score":0.17339999973773956,"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/T10639","display_name":"Advanced Software Engineering Methodologies","score":0.17339999973773956,"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/T11241","display_name":"Advanced Malware Detection Techniques","score":0.11420000344514847,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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":0.07649999856948853,"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/adaptation","display_name":"Adaptation (eye)","score":0.7226999998092651},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.6668999791145325},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.6031000018119812},{"id":"https://openalex.org/keywords/action","display_name":"Action (physics)","score":0.4510999917984009},{"id":"https://openalex.org/keywords/mobile-apps","display_name":"Mobile apps","score":0.43950000405311584},{"id":"https://openalex.org/keywords/cover","display_name":"Cover (algebra)","score":0.41280001401901245},{"id":"https://openalex.org/keywords/mobile-device","display_name":"Mobile device","score":0.40529999136924744},{"id":"https://openalex.org/keywords/trajectory","display_name":"Trajectory","score":0.37860000133514404}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7425000071525574},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.7226999998092651},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.6668999791145325},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.6031000018119812},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.48170000314712524},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.4510999917984009},{"id":"https://openalex.org/C2988145974","wikidata":"https://www.wikidata.org/wiki/Q620615","display_name":"Mobile apps","level":2,"score":0.43950000405311584},{"id":"https://openalex.org/C2780428219","wikidata":"https://www.wikidata.org/wiki/Q16952335","display_name":"Cover (algebra)","level":2,"score":0.41280001401901245},{"id":"https://openalex.org/C186967261","wikidata":"https://www.wikidata.org/wiki/Q5082128","display_name":"Mobile device","level":2,"score":0.40529999136924744},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.37860000133514404},{"id":"https://openalex.org/C144543869","wikidata":"https://www.wikidata.org/wiki/Q2738570","display_name":"Mobile computing","level":2,"score":0.35260000824928284},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.31850001215934753},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.27900001406669617},{"id":"https://openalex.org/C2779436431","wikidata":"https://www.wikidata.org/wiki/Q30672407","display_name":"Policy learning","level":2,"score":0.2712000012397766},{"id":"https://openalex.org/C52970973","wikidata":"https://www.wikidata.org/wiki/Q2497134","display_name":"Adaptive system","level":2,"score":0.25839999318122864},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.25769999623298645},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.25589999556541443},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.25130000710487366},{"id":"https://openalex.org/C149672232","wikidata":"https://www.wikidata.org/wiki/Q337048","display_name":"Adaptive optimization","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.19930","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19930","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.19930","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19930","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"MLLM-based":[0],"mobile":[1,24,90,106,177],"GUI":[2,43,91,178],"agents":[3],"have":[4],"made":[5],"substantial":[6],"progress":[7],"in":[8,104,180],"UI":[9],"understanding":[10],"and":[11,30,60,70,101,109,122,164,185],"action":[12],"execution,":[13,59],"but":[14],"adapting":[15],"them":[16],"to":[17,32,76,140,146],"real":[18,105],"target":[19],"apps":[20,25],"remains":[21],"costly":[22],"because":[23],"are":[26,74],"numerous,":[27],"frequently":[28],"updated,":[29],"hard":[31,75],"cover":[33],"with":[34],"human-written":[35],"tasks,":[36],"demonstrations,":[37],"or":[38],"reward":[39],"labels.":[40],"Existing":[41],"annotation-free":[42,86,134],"learning":[44],"reduces":[45],"manual":[46],"supervision,":[47],"yet":[48],"lacks":[49],"a":[50],"unified":[51],"substrate":[52],"connecting":[53],"target-app":[54],"exploration,":[55],"curriculum":[56],"mining,":[57],"rollout":[58,102],"feedback,":[61,121],"while":[62],"policy":[63],"optimization":[64],"often":[65],"relies":[66],"on":[67,143,162,167],"isolated":[68],"rollouts":[69],"coarse":[71],"rewards":[72],"that":[73],"convert":[77],"into":[78,125],"reliable":[79],"improvement":[80],"signals.":[81],"We":[82],"present":[83],"MobileForge,":[84],"an":[85],"adaptation":[87,135],"system":[88],"for":[89],"agents.":[92],"MobileForge":[93,137],"consists":[94],"of":[95],"MobileGym,":[96],"which":[97,115],"grounds":[98],"task":[99],"generation":[100],"evaluation":[103],"app":[107],"interaction,":[108],"Hierarchical":[110],"Feedback-Guided":[111],"Policy":[112],"Optimization":[113],"(HiFPO),":[114],"turns":[116],"trajectory":[117],"outcomes,":[118],"step-level":[119,127],"process":[120],"corrective":[123],"hints":[124],"hint-contextualized":[126],"GRPO":[128],"updates.":[129],"Using":[130],"only":[131],"automatically":[132],"generated":[133],"data,":[136,184],"adapts":[138],"Qwen3-VL-8B":[139],"67.2%":[141],"Pass@3":[142,161],"AndroidWorld,":[144],"close":[145],"the":[147,168,174],"closed-data":[148],"GUI-specialized":[149],"GUI-Owl-1.5-8B":[150],"base":[151],"model":[152],"at":[153,191],"69.0%.":[154],"The":[155],"MobileForge-adapted":[156],"ForgeOwl-8B":[157],"further":[158],"reaches":[159],"77.6%":[160],"AndroidWorld":[163],"41.0%":[165],"success":[166],"out-of-domain":[169],"MobileWorld":[170],"GUI-only":[171],"split,":[172],"establishing":[173],"strongest":[175],"open-data":[176],"agent":[179],"our":[181],"evaluation.":[182],"Code,":[183],"trained":[186],"models":[187],"will":[188],"be":[189],"released":[190],"https://mobile-forge.github.io/.":[192]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-20T00:00:00"}