{"id":"https://openalex.org/W7157292097","doi":"https://doi.org/10.48550/arxiv.2604.24182","title":"$M^2$-VLA: Boosting Vision-Language Models for Generalizable Manipulation via Layer Mixture and Meta-Skills","display_name":"$M^2$-VLA: Boosting Vision-Language Models for Generalizable Manipulation via Layer Mixture and Meta-Skills","publication_year":2026,"publication_date":"2026-04-27","ids":{"openalex":"https://openalex.org/W7157292097","doi":"https://doi.org/10.48550/arxiv.2604.24182"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.24182","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24182","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.24182","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134819970","display_name":"Siyao Xiao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiao, Siyao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134806701","display_name":"Yuhong Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Yuhong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134817145","display_name":"Zhifang Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zhifang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134796001","display_name":"Zihan Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Zihan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134796477","display_name":"Jingye Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Jingye","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134761063","display_name":"Sinwai Choo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Choo, Sinwai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134758895","display_name":"Dake Zhong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhong, Dake","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134771349","display_name":"Mengzhe Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Mengzhe","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134809001","display_name":"Xiao Lin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Xiao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134798405","display_name":"Xianfeng Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Xianfeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134749549","display_name":"Jia Jia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jia, Jia","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134758637","display_name":"Haoqian Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Haoqian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":12,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.5687999725341797,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.5687999725341797,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10653","display_name":"Robot Manipulation and Learning","score":0.15559999644756317,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.08129999786615372,"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/boosting","display_name":"Boosting (machine learning)","score":0.671500027179718},{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.6355000138282776},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.6287000179290771},{"id":"https://openalex.org/keywords/bridge","display_name":"Bridge (graph theory)","score":0.49399998784065247},{"id":"https://openalex.org/keywords/layer","display_name":"Layer (electronics)","score":0.43130001425743103},{"id":"https://openalex.org/keywords/robot","display_name":"Robot","score":0.37209999561309814},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.3562000095844269}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7484999895095825},{"id":"https://openalex.org/C46686674","wikidata":"https://www.wikidata.org/wiki/Q466303","display_name":"Boosting (machine learning)","level":2,"score":0.671500027179718},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.6355000138282776},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.6287000179290771},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6212999820709229},{"id":"https://openalex.org/C100776233","wikidata":"https://www.wikidata.org/wiki/Q2532492","display_name":"Bridge (graph theory)","level":2,"score":0.49399998784065247},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4490000009536743},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.43130001425743103},{"id":"https://openalex.org/C90509273","wikidata":"https://www.wikidata.org/wiki/Q11012","display_name":"Robot","level":2,"score":0.37209999561309814},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.3562000095844269},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.34139999747276306},{"id":"https://openalex.org/C34413123","wikidata":"https://www.wikidata.org/wiki/Q170978","display_name":"Robotics","level":3,"score":0.30379998683929443},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.29739999771118164},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.2872999906539917},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.28619998693466187},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.271699994802475},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.2632000148296356}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.24182","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24182","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.24182","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24182","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[{"display_name":"Quality Education","score":0.7479234933853149,"id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Current":[0],"Vision-Language-Action":[1],"(VLA)":[2],"models":[3,153],"predominantly":[4],"rely":[5],"on":[6],"end-to-end":[7],"fine-tuning.":[8],"While":[9],"effective,":[10],"this":[11],"paradigm":[12],"compromises":[13],"the":[14,59,62,69,80,126,137,143],"inherent":[15],"generalization":[16,132],"capabilities":[17,140],"of":[18,66,72,82,128,145],"Vision-Language":[19],"Models":[20],"(VLMs)":[21],"and":[22,68,122,133,141,151],"incurs":[23],"catastrophic":[24],"forgetting.":[25],"To":[26,75],"address":[27],"these":[28],"limitations,":[29],"we":[30,78,105],"propose":[31,106],"$M^2$-VLA,":[32],"which":[33],"demonstrates":[34],"that":[35,86,112],"a":[36,44,54,107],"generalized":[37],"VLM":[38],"is":[39],"able":[40],"to":[41,57,96],"serve":[42],"as":[43],"powerful":[45],"backbone":[46],"for":[47],"robotic":[48,73],"manipulation":[49],"directly.":[50],"However,":[51],"it":[52],"remains":[53],"key":[55,147],"challenge":[56],"bridge":[58],"gap":[60],"between":[61],"high-level":[63],"semantic":[64,93],"understanding":[65],"VLMs":[67],"precise":[70],"requirements":[71],"control.":[74],"overcome":[76],"this,":[77],"introduce":[79],"Mixture":[81],"Layers":[83],"(MoL)":[84],"strategy":[85],"selectively":[87],"extracts":[88],"task-critical":[89],"information":[90],"from":[91],"dense":[92],"features.":[94],"Furthermore,":[95,131],"facilitate":[97],"efficient":[98],"trajectory":[99],"learning":[100],"under":[101],"constrained":[102],"model":[103],"capacity,":[104],"Meta":[108],"Skill":[109],"Module":[110],"(MSM)":[111],"integrates":[113],"strong":[114],"inductive":[115],"biases.":[116],"Extensive":[117],"experiments":[118],"in":[119],"both":[120],"simulated":[121],"real-world":[123],"environments":[124],"demonstrate":[125],"effectiveness":[127],"our":[129],"approach.":[130],"ablation":[134],"studies":[135],"validate":[136],"architecture's":[138],"zero-shot":[139],"confirm":[142],"contribution":[144],"each":[146],"component.":[148],"Our":[149],"code":[150],"pre-trained":[152],"will":[154],"be":[155],"made":[156],"publicly":[157],"available.":[158]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-04-29T00:00:00"}