{"id":"https://openalex.org/W7138382548","doi":"https://doi.org/10.48550/arxiv.2603.14972","title":"Learning from Mistakes: Post-Training for Driving VLA with Takeover Data","display_name":"Learning from Mistakes: Post-Training for Driving VLA with Takeover Data","publication_year":2026,"publication_date":"2026-03-16","ids":{"openalex":"https://openalex.org/W7138382548","doi":"https://doi.org/10.48550/arxiv.2603.14972"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.14972","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14972","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.2603.14972","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129641353","display_name":"Yinfeng Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Yinfeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129740307","display_name":"Deqing Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Deqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129655390","display_name":"Qichao Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Qichao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129723689","display_name":"Yupeng Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Yupeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129738323","display_name":"Haochen Tian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tian, Haochen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129697704","display_name":"Guang Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Guang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129701305","display_name":"Hangjun Ye","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ye, Hangjun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129713475","display_name":"Long Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Long","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073089333","display_name":"Da\u2010Wei Ding","orcid":"https://orcid.org/0000-0003-1201-7785"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ding, Da-Wei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5129752531","display_name":"Dongbin Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Dongbin","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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.2971000075340271,"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":0.2971000075340271,"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/T11099","display_name":"Autonomous Vehicle Technology and Safety","score":0.22220000624656677,"subfield":{"id":"https://openalex.org/subfields/2203","display_name":"Automotive 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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.06719999760389328,"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/key","display_name":"Key (lock)","score":0.6722999811172485},{"id":"https://openalex.org/keywords/margin","display_name":"Margin (machine learning)","score":0.6664999723434448},{"id":"https://openalex.org/keywords/mindset","display_name":"Mindset","score":0.6539000272750854},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.5716999769210815},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.5485000014305115},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.5259000062942505},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.5159000158309937},{"id":"https://openalex.org/keywords/function","display_name":"Function (biology)","score":0.30480000376701355}],"concepts":[{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.6722999811172485},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.6664999723434448},{"id":"https://openalex.org/C2778491294","wikidata":"https://www.wikidata.org/wiki/Q1339824","display_name":"Mindset","level":2,"score":0.6539000272750854},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6292999982833862},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5716999769210815},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.5485000014305115},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5259000062942505},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.5159000158309937},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.460099995136261},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.4108999967575073},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.31380000710487366},{"id":"https://openalex.org/C42475967","wikidata":"https://www.wikidata.org/wiki/Q194292","display_name":"Operations research","level":1,"score":0.3070000112056732},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.30480000376701355},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.30079999566078186},{"id":"https://openalex.org/C77967617","wikidata":"https://www.wikidata.org/wiki/Q4677561","display_name":"Active learning (machine learning)","level":2,"score":0.29760000109672546},{"id":"https://openalex.org/C2779436431","wikidata":"https://www.wikidata.org/wiki/Q30672407","display_name":"Policy learning","level":2,"score":0.2759999930858612},{"id":"https://openalex.org/C7493553","wikidata":"https://www.wikidata.org/wiki/Q1520777","display_name":"Certainty","level":2,"score":0.27399998903274536},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.2711000144481659},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.27090001106262207},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.267300009727478},{"id":"https://openalex.org/C127729010","wikidata":"https://www.wikidata.org/wiki/Q60165","display_name":"Dynamic inconsistency","level":2,"score":0.2615000009536743},{"id":"https://openalex.org/C29122968","wikidata":"https://www.wikidata.org/wiki/Q1414816","display_name":"Incentive","level":2,"score":0.2572999894618988}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.14972","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14972","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.2603.14972","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14972","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":[{"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions","score":0.6426156163215637}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Current":[0],"Vision-Language-Action":[1],"(VLA)":[2],"paradigms":[3],"in":[4,115,143,174,188],"end-to-end":[5],"autonomous":[6],"driving":[7,175],"rely":[8],"on":[9,155],"offline":[10],"training":[11],"from":[12,42,103],"static":[13],"datasets,":[14],"leaving":[15],"them":[16],"vulnerable":[17],"to":[18,27,48,56,101,113],"distribution":[19],"shift.":[20],"Recent":[21],"post-training":[22,81],"methods":[23],"use":[24],"takeover":[25,37,53,145],"data":[26],"mitigate":[28],"this":[29,73],"by":[30,172,184],"augmenting":[31],"the":[32,49,52,99,109,156,167],"dataset":[33],"with":[34,58,177],"high-quality":[35],"expert":[36],"samples,":[38],"yet":[39],"they":[40],"suffer":[41],"two":[43,88],"key":[44],"limitations:":[45],"supervision":[46],"restricted":[47],"period":[50],"after":[51],"moments":[54],"leads":[55],"policies":[57],"limited":[59],"safety":[60,130,180],"margins,":[61],"and":[62,127],"passive":[63],"preference":[64,152],"optimization":[65],"lacks":[66],"active":[67,148],"exploration":[68,149],"for":[69],"optimal":[70],"performance.":[71],"In":[72],"paper,":[74],"we":[75,92,118,133],"propose":[76,134],"TakeVLA,":[77],"a":[78,120,137],"novel":[79],"VLA":[80,100,169],"framework":[82],"that":[83,123,141,160],"overcomes":[84],"these":[85],"shortcomings":[86],"through":[87],"complementary":[89],"innovations.":[90],"First,":[91],"introduce":[93],"pre-takeover":[94],"language":[95],"supervision,":[96],"which":[97],"allows":[98],"learn":[102],"mistakes":[104],"proactively.":[105],"By":[106],"explicitly":[107],"teaching":[108],"model":[110],"about":[111],"what":[112],"do":[114],"error-prone":[116],"situations,":[117],"cultivate":[119],"precautionary":[121],"mindset":[122],"anticipates":[124],"hazards":[125],"early":[126],"substantially":[128],"enlarges":[129],"margins.":[131],"Second,":[132],"Scenario":[135],"Dreaming,":[136],"reinforcement":[138],"fine-tuning":[139],"paradigm":[140],"operates":[142],"reconstruceted":[144],"scenarios,":[146],"encouraging":[147],"beyond":[150],"mere":[151],"fitting.":[153],"Experiments":[154],"Bench2Drive":[157],"benchmark":[158],"demonstrate":[159],"TakeVLA":[161],"achieves":[162],"state-of-the-art":[163],"closed-loop":[164],"performance,":[165],"surpassing":[166],"strong":[168],"baseline":[170],"SimLingo":[171],"4.93":[173],"score,":[176],"an":[178,185],"enhanced":[179],"margin":[181],"as":[182],"evidenced":[183],"11.76%":[186],"increase":[187],"average":[189],"TTC.":[190]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-18T00:00:00"}