{"id":"https://openalex.org/W7131397440","doi":"https://doi.org/10.48550/arxiv.2602.20332","title":"No One Size Fits All: QueryBandits for Hallucination Mitigation","display_name":"No One Size Fits All: QueryBandits for Hallucination Mitigation","publication_year":2026,"publication_date":"2026-02-23","ids":{"openalex":"https://openalex.org/W7131397440","doi":"https://doi.org/10.48550/arxiv.2602.20332"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.20332","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","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":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5084455404","display_name":"Nicole Cho","orcid":"https://orcid.org/0009-0006-9007-3255"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Cho, Nicole","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074837370","display_name":"William Watson","orcid":"https://orcid.org/0000-0001-5516-262X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Watson, William","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126787518","display_name":"Alec Koppel","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Koppel, Alec","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081450666","display_name":"Sumitra Ganesh","orcid":"https://orcid.org/0000-0003-1695-8574"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ganesh, Sumitra","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5126842496","display_name":"Manuela Veloso","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Veloso, Manuela","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5084455404"],"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.31949999928474426,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.31949999928474426,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.07829999923706055,"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/T10028","display_name":"Topic Modeling","score":0.0689999982714653,"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/paraphrase","display_name":"Paraphrase","score":0.8134999871253967},{"id":"https://openalex.org/keywords/regret","display_name":"Regret","score":0.6559000015258789},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.6093999743461609},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.5875999927520752},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.5644000172615051},{"id":"https://openalex.org/keywords/retraining","display_name":"Retraining","score":0.5370000004768372},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4560999870300293},{"id":"https://openalex.org/keywords/work","display_name":"Work (physics)","score":0.42320001125335693},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.37070000171661377}],"concepts":[{"id":"https://openalex.org/C2780922921","wikidata":"https://www.wikidata.org/wiki/Q255189","display_name":"Paraphrase","level":2,"score":0.8134999871253967},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7239999771118164},{"id":"https://openalex.org/C50817715","wikidata":"https://www.wikidata.org/wiki/Q79895177","display_name":"Regret","level":2,"score":0.6559000015258789},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6101999878883362},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.6093999743461609},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.5875999927520752},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.5644000172615051},{"id":"https://openalex.org/C2778712577","wikidata":"https://www.wikidata.org/wiki/Q3505966","display_name":"Retraining","level":2,"score":0.5370000004768372},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5205000042915344},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4560999870300293},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.42320001125335693},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.37070000171661377},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.3411000072956085},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.33980000019073486},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.33820000290870667},{"id":"https://openalex.org/C2777530160","wikidata":"https://www.wikidata.org/wiki/Q41796","display_name":"Sentence","level":2,"score":0.32749998569488525},{"id":"https://openalex.org/C2776502983","wikidata":"https://www.wikidata.org/wiki/Q690182","display_name":"Contrast (vision)","level":2,"score":0.32190001010894775},{"id":"https://openalex.org/C73602740","wikidata":"https://www.wikidata.org/wiki/Q7795822","display_name":"Thompson sampling","level":3,"score":0.3174000084400177},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.29809999465942383},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.29649999737739563},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.28540000319480896},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.27639999985694885},{"id":"https://openalex.org/C169258074","wikidata":"https://www.wikidata.org/wiki/Q245748","display_name":"Random forest","level":2,"score":0.2653000056743622},{"id":"https://openalex.org/C4679612","wikidata":"https://www.wikidata.org/wiki/Q866298","display_name":"Aggregate (composite)","level":2,"score":0.263700008392334},{"id":"https://openalex.org/C149629883","wikidata":"https://www.wikidata.org/wiki/Q660926","display_name":"Fraction (chemistry)","level":2,"score":0.262800008058548},{"id":"https://openalex.org/C2986087404","wikidata":"https://www.wikidata.org/wiki/Q15946010","display_name":"Online learning","level":2,"score":0.25940001010894775},{"id":"https://openalex.org/C155108698","wikidata":"https://www.wikidata.org/wiki/Q1231081","display_name":"Randomized experiment","level":2,"score":0.2531000077724457}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.20332","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},{"id":"doi:10.48550/arxiv.2602.20332","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.20332","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":"article"}],"best_oa_location":{"id":"pmh:doi:10.48550/arxiv.2602.20332","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions","score":0.4176178574562073}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Advanced":[0],"reasoning":[1],"capabilities":[2],"in":[3,35,49,129],"Large":[4],"Language":[5],"Models":[6],"(LLMs)":[7],"have":[8],"led":[9],"to":[10,64],"more":[11],"frequent":[12],"hallucinations;":[13],"yet":[14],"most":[15],"mitigation":[16],"work":[17],"focuses":[18],"on":[19,33],"open-source":[20],"models":[21,37,48,192],"for":[22,144,197],"post-hoc":[23],"detection":[24],"and":[25,75,97,108,193],"parameter":[26],"editing.":[27],"The":[28],"dearth":[29],"of":[30,47],"studies":[31],"focusing":[32],"hallucinations":[34],"closed-source":[36,191],"is":[38,138],"especially":[39],"concerning,":[40],"as":[41],"they":[42],"constitute":[43],"the":[44,66,195],"vast":[45],"majority":[46],"institutional":[50],"deployments.":[51],"We":[52,147],"introduce":[53],"QueryBandits,":[54],"a":[55,94],"model-agnostic":[56],"contextual":[57,113],"bandit":[58],"framework":[59],"that":[60,136,150,161],"adaptively":[61],"learns":[62],"online":[63,172],"select":[65],"optimal":[67,143],"query-rewrite":[68],"strategy":[69],"by":[70,106],"leveraging":[71],"an":[72,89,162,171],"empirically":[73],"validated":[74],"calibrated":[76],"reward":[77],"function.":[78],"Across":[79],"16":[80],"QA":[81],"scenarios,":[82],"our":[83,134],"top":[84],"QueryBandit":[85],"(Thompson":[86],"Sampling)":[87],"achieves":[88],"87.5%":[90],"win":[91],"rate":[92],"over":[93,174],"No-Rewrite":[95],"baseline":[96],"outperforms":[98],"zero-shot":[99],"static":[100,152],"policies":[101,153],"(e.g.,":[102],"Paraphrase":[103],"or":[104,199],"Expand)":[105],"42.6%":[107],"60.3%,":[109],"respectively.":[110],"Moreover,":[111],"all":[112,119,145],"bandits":[114,117],"outperform":[115],"vanilla":[116],"across":[118],"datasets,":[120],"with":[121,126,177,190],"higher":[122,155],"feature":[123],"variance":[124,128],"coinciding":[125],"greater":[127],"arm":[130],"selection.":[131],"This":[132],"substantiates":[133],"finding":[135],"there":[137],"no":[139],"single":[140],"rewrite":[141],"policy":[142,165,173],"queries.":[146],"also":[148],"discover":[149],"certain":[151],"incur":[154],"cumulative":[156],"regret":[157],"than":[158],"No-Rewrite,":[159],"indicating":[160],"inflexible":[163],"query-rewriting":[164],"can":[166,179],"worsen":[167],"hallucinations.":[168],"Thus,":[169],"learning":[170],"semantic":[175],"features":[176],"QueryBandits":[178],"shift":[180],"model":[181],"behavior":[182],"purely":[183],"through":[184],"forward-pass":[185],"mechanisms,":[186],"enabling":[187],"its":[188],"use":[189],"bypassing":[194],"need":[196],"retraining":[198],"gradient-based":[200],"adaptation.":[201]},"counts_by_year":[],"updated_date":"2026-04-04T16:13:02.066488","created_date":"2026-02-26T00:00:00"}