{"id":"https://openalex.org/W7160315562","doi":"https://doi.org/10.48550/arxiv.2605.01529","title":"Good in Bad (GiB): Sifting Through End-user Demonstrations for Learning a Better Policy","display_name":"Good in Bad (GiB): Sifting Through End-user Demonstrations for Learning a Better Policy","publication_year":2026,"publication_date":"2026-05-02","ids":{"openalex":"https://openalex.org/W7160315562","doi":"https://doi.org/10.48550/arxiv.2605.01529"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.01529","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.01529","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.2605.01529","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5086743894","display_name":"Noushad Sojib","orcid":"https://orcid.org/0009-0008-7301-4741"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sojib, Noushad","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135296177","display_name":"Momotaz Begum","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ghattas, Ola","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Begum, Momotaz","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Begum, Momotaz","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":3,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.6818000078201294,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.6818000078201294,"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/T10653","display_name":"Robot Manipulation and Learning","score":0.12229999899864197,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.024900000542402267,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.6010000109672546},{"id":"https://openalex.org/keywords/mahalanobis-distance","display_name":"Mahalanobis distance","score":0.512499988079071},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.4747999906539917},{"id":"https://openalex.org/keywords/imitation","display_name":"Imitation","score":0.4677000045776367},{"id":"https://openalex.org/keywords/policy-learning","display_name":"Policy learning","score":0.40529999136924744},{"id":"https://openalex.org/keywords/complement","display_name":"Complement (music)","score":0.38190001249313354},{"id":"https://openalex.org/keywords/robot","display_name":"Robot","score":0.37869998812675476},{"id":"https://openalex.org/keywords/staring","display_name":"Staring","score":0.3668999969959259},{"id":"https://openalex.org/keywords/feature-learning","display_name":"Feature learning","score":0.3497999906539917}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6873000264167786},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6229000091552734},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.6010000109672546},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5138999819755554},{"id":"https://openalex.org/C1921717","wikidata":"https://www.wikidata.org/wiki/Q1334846","display_name":"Mahalanobis distance","level":2,"score":0.512499988079071},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4747999906539917},{"id":"https://openalex.org/C126388530","wikidata":"https://www.wikidata.org/wiki/Q1131737","display_name":"Imitation","level":2,"score":0.4677000045776367},{"id":"https://openalex.org/C2779436431","wikidata":"https://www.wikidata.org/wiki/Q30672407","display_name":"Policy learning","level":2,"score":0.40529999136924744},{"id":"https://openalex.org/C112313634","wikidata":"https://www.wikidata.org/wiki/Q7886648","display_name":"Complement (music)","level":5,"score":0.38190001249313354},{"id":"https://openalex.org/C90509273","wikidata":"https://www.wikidata.org/wiki/Q11012","display_name":"Robot","level":2,"score":0.37869998812675476},{"id":"https://openalex.org/C2778949103","wikidata":"https://www.wikidata.org/wiki/Q600717","display_name":"Staring","level":2,"score":0.3668999969959259},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.3497999906539917},{"id":"https://openalex.org/C2778827112","wikidata":"https://www.wikidata.org/wiki/Q22245680","display_name":"Feature engineering","level":3,"score":0.3472999930381775},{"id":"https://openalex.org/C77967617","wikidata":"https://www.wikidata.org/wiki/Q4677561","display_name":"Active learning (machine learning)","level":2,"score":0.3271999955177307},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.32690000534057617},{"id":"https://openalex.org/C190839683","wikidata":"https://www.wikidata.org/wiki/Q2448197","display_name":"Train","level":2,"score":0.30379998683929443},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.2782999873161316},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.2694000005722046},{"id":"https://openalex.org/C24138899","wikidata":"https://www.wikidata.org/wiki/Q17141258","display_name":"Instance-based learning","level":3,"score":0.2662000060081482},{"id":"https://openalex.org/C188888258","wikidata":"https://www.wikidata.org/wiki/Q7353390","display_name":"Robot learning","level":4,"score":0.265500009059906},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.26330000162124634},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2597000002861023},{"id":"https://openalex.org/C66905080","wikidata":"https://www.wikidata.org/wiki/Q17005494","display_name":"Binary classification","level":3,"score":0.2563999891281128},{"id":"https://openalex.org/C51167844","wikidata":"https://www.wikidata.org/wiki/Q4422623","display_name":"Latent variable","level":2,"score":0.2549000084400177},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2542000114917755}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.01529","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.01529","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.2605.01529","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.01529","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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Imitation":[0],"learning":[1,19,42,99,168],"offers":[2],"a":[3,109],"promising":[4],"framework":[5],"for":[6],"enabling":[7],"robots":[8],"to":[9,23,57,101,112,120,143],"acquire":[10],"diverse":[11],"skills":[12],"from":[13,32,43,169],"human":[14,171],"users.":[15],"However,":[16],"most":[17],"imitation":[18],"algorithms":[20],"assume":[21],"access":[22],"high-quality":[24,87,136],"demonstrations":[25,36,45,55,84],"an":[26,74],"unrealistic":[27],"expectation":[28],"when":[29,167],"collecting":[30],"data":[31,91],"non-expert":[33],"users,":[34],"whose":[35],"often":[37],"contain":[38],"inadvertent":[39],"errors.":[40],"Naively":[41],"such":[44],"can":[46,92],"result":[47],"in":[48,64,156],"unsafe":[49],"policy":[50,98,165],"behavior,":[51],"while":[52,85],"discarding":[53],"entire":[54],"due":[56],"occasional":[58],"mistakes":[59],"wastes":[60],"valuable":[61],"data,":[62],"especially":[63],"low-data":[65],"settings.":[66],"In":[67],"this":[68],"work,":[69],"we":[70],"introduce":[71],"GiB":[72,106,151],"(Good-in-Bad),":[73],"algorithm":[75,100],"that":[76],"automatically":[77],"identifies":[78],"and":[79,116,138,145,159],"discards":[80],"erroneous":[81],"subtasks":[82],"within":[83],"preserving":[86],"subtasks.":[88,148],"The":[89],"filtered":[90],"then":[93,129],"be":[94],"used":[95],"by":[96],"any":[97],"train":[102],"more":[103],"robust":[104],"policies.":[105],"first":[107],"trains":[108],"self-supervised":[110],"model":[111],"learn":[113],"latent":[114,132],"features":[115],"assigns":[117],"binary":[118],"weights":[119],"label":[121],"each":[122],"demonstration":[123],"as":[124],"good":[125],"or":[126],"bad.":[127],"It":[128],"models":[130],"the":[131,140,153],"feature":[133],"distribution":[134],"of":[135],"segments":[137],"uses":[139],"Mahalanobis":[141],"distance":[142],"detect":[144],"evaluate":[146],"poor-quality":[147],"We":[149],"validate":[150],"on":[152],"Franka":[154],"robot":[155],"both":[157],"simulated":[158],"real-world":[160],"multi-step":[161],"tasks,":[162],"demonstrating":[163],"improved":[164],"performance":[166],"mixed-quality":[170],"demonstrations.":[172]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-05-06T00:00:00"}