{"id":"https://openalex.org/W7163038009","doi":"https://doi.org/10.48550/arxiv.2605.30660","title":"BOKBO (Best of K Bad Options): Calibrated Abstention for VLA Policies","display_name":"BOKBO (Best of K Bad Options): Calibrated Abstention for VLA Policies","publication_year":2026,"publication_date":"2026-05-28","ids":{"openalex":"https://openalex.org/W7163038009","doi":"https://doi.org/10.48550/arxiv.2605.30660"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.30660","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30660","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.2605.30660","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5130283294","display_name":"Anya Singh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Singh, Anya","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137608037","display_name":"Cabrel Happi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Happi, Cabrel","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137546196","display_name":"Jai Relan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Relan, Jai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137607477","display_name":"Varun Nair","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nair, Varun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137601616","display_name":"Vidyut Baradwaj","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Baradwaj, Vidyut","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":0.6046000123023987,"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":0.6046000123023987,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.08060000091791153,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.06239999830722809,"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/replicate","display_name":"Replicate","score":0.7342000007629395},{"id":"https://openalex.org/keywords/spurious-relationship","display_name":"Spurious relationship","score":0.637499988079071},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5270000100135803},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.40540000796318054},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.40230000019073486},{"id":"https://openalex.org/keywords/proxy","display_name":"Proxy (statistics)","score":0.3889000117778778},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.34060001373291016},{"id":"https://openalex.org/keywords/conditional-independence","display_name":"Conditional independence","score":0.32519999146461487}],"concepts":[{"id":"https://openalex.org/C2781162219","wikidata":"https://www.wikidata.org/wiki/Q26250693","display_name":"Replicate","level":2,"score":0.7342000007629395},{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.637499988079071},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.5388000011444092},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5270000100135803},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4424000084400177},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4066999852657318},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.40540000796318054},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.40230000019073486},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.3977000117301941},{"id":"https://openalex.org/C2780148112","wikidata":"https://www.wikidata.org/wiki/Q1432581","display_name":"Proxy (statistics)","level":2,"score":0.3889000117778778},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.34060001373291016},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.33250001072883606},{"id":"https://openalex.org/C79772020","wikidata":"https://www.wikidata.org/wiki/Q5159264","display_name":"Conditional independence","level":2,"score":0.32519999146461487},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.31439998745918274},{"id":"https://openalex.org/C130440534","wikidata":"https://www.wikidata.org/wiki/Q14946528","display_name":"Conflation","level":2,"score":0.29820001125335693},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.29019999504089355},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.2854999899864197},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.28290000557899475},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.28040000796318054},{"id":"https://openalex.org/C43555835","wikidata":"https://www.wikidata.org/wiki/Q2300258","display_name":"Conditional probability distribution","level":2,"score":0.2775999903678894},{"id":"https://openalex.org/C87007009","wikidata":"https://www.wikidata.org/wiki/Q210832","display_name":"Statistical hypothesis testing","level":2,"score":0.27410000562667847},{"id":"https://openalex.org/C44492722","wikidata":"https://www.wikidata.org/wiki/Q327069","display_name":"Conditional probability","level":2,"score":0.26919999718666077},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2667999863624573},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.2651999890804291},{"id":"https://openalex.org/C2778355321","wikidata":"https://www.wikidata.org/wiki/Q17079427","display_name":"Identity (music)","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.26030001044273376},{"id":"https://openalex.org/C38935604","wikidata":"https://www.wikidata.org/wiki/Q4330363","display_name":"Stylized fact","level":2,"score":0.250900000333786}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.30660","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30660","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.2605.30660","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30660","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":[{"display_name":"Peace, Justice and strong institutions","id":"https://metadata.un.org/sdg/16","score":0.8070440292358398}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Test-time":[0],"scaling":[1],"for":[2,48],"vision-language-action":[3],"(VLA)":[4],"policies,":[5],"methods":[6],"such":[7],"as":[8,211],"RoboMonkey,":[9],"SEAL,":[10],"MG-Select,":[11],"and":[12,21,63,96,130,136,151,178,215,224],"V-GPS,":[13],"samples":[14],"K":[15,27],"candidate":[16],"action":[17,36],"chunks":[18],"at":[19,100,109,157],"inference":[20],"executes":[22,33],"the":[23,31,43,68,72,76,92,103,110,119,124,132,165,185],"verifier-best.":[24],"When":[25],"all":[26],"candidates":[28],"are":[29,196],"unsafe,":[30],"system":[32],"a":[34,82,212,226],"violating":[35],"with":[37,67,102,113,163,175,240],"no":[38],"warning.":[39],"We":[40,59,117,221],"propose":[41],"BOKBO,":[42],"first":[44],"conformal":[45],"abstention":[46],"layer":[47],"K-sample":[49,97],"VLA":[50],"inference,":[51],"providing":[52],"finite-sample":[53],"distribution-free":[54],"guarantees":[55],"on":[56,75,147,161,170,206,209],"executed-violation":[57],"rate.":[58],"provide":[60],"both":[61],"global":[62],"per-task":[64,69,187],"(Mondrian)":[65],"variants,":[66],"variant":[70],"closing":[71],"conditional":[73,166,188],"gap":[74],"hardest":[77],"tasks.":[78],"Our":[79],"analysis":[80,125],"exposes":[81],"structural":[83],"failure":[84,133],"of":[85,172],"policy-internal":[86],"nonconformity":[87],"scores":[88],"under":[89,126,139],"perturbation-based":[90],"K-sampling:":[91],"base-policy":[93],"confidence":[94],"proxy":[95],"disagreement":[98],"correlate":[99],"0.98":[101],"action-noise":[104],"hyperparameter":[105],"$\u03c3$,":[106],"while":[107],"correlating":[108],"noise":[111,205],"floor":[112],"actual":[114],"safety":[115],"violations.":[116],"test":[118],"failure's":[120],"scope":[121],"by":[122,246],"replicating":[123],"token-level":[127],"temperature":[128],"sampling":[129],"find":[131],"is":[134],"mechanism-specific":[135],"partially":[137],"mitigated":[138],"policy-stochasticity-based":[140],"sampling.":[141],"A":[142],"learned":[143],"violation":[144,244],"predictor":[145],"conditioned":[146],"semantic":[148],"visual":[149],"features":[150],"task":[152,181],"identity":[153],"supports":[154],"tight":[155],"calibration:":[156],"$\u03b5$":[158],"=":[159],"0.05":[160],"libero_object_temp_x0.1":[162],"OpenVLA-OFT,":[164],"CRC":[167],"bound":[168],"holds":[169],"86%":[171],"bootstrap":[173,204],"splits":[174],"78%":[176],"coverage":[177],"70%":[179],"net":[180],"success.":[182],"Mondrian-BOKBO":[183],"raises":[184],"minimum":[186],"hold":[189,208],"fraction":[190],"from":[191],"0.71":[192],"to":[193],"0.93.":[194],"Results":[195],"stable":[197],"across":[198],"5":[199],"training":[200],"seeds,":[201],"replicate":[202],"within":[203],"$\u03c0_0$-FAST,":[207],"libero_spatial_temp_x0.1":[210],"co-equal":[213],"benchmark,":[214],"survive":[216],"four":[217],"within-suite":[218],"distribution":[219],"shifts.":[220],"additionally":[222],"identify":[223],"correct":[225],"methodological":[227],"pitfall:":[228],"globally-set":[229],"force":[230],"thresholds":[231],"well":[232],"below":[233],"expert-typical":[234],"manipulation":[235],"forces":[236],"conflate":[237],"unsafe":[238],"behavior":[239],"normal":[241],"manipulation,":[242],"inflating":[243],"rates":[245],"$5\\times$.":[247]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-02T00:00:00"}