{"id":"https://openalex.org/W7161141142","doi":"https://doi.org/10.48550/arxiv.2605.13369","title":"Query-Conditioned Test-Time Self-Training for Large Language Models","display_name":"Query-Conditioned Test-Time Self-Training for Large Language Models","publication_year":2026,"publication_date":"2026-05-13","ids":{"openalex":"https://openalex.org/W7161141142","doi":"https://doi.org/10.48550/arxiv.2605.13369"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.13369","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.13369","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.2605.13369","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5036706045","display_name":"Chaehee Song","orcid":"https://orcid.org/0000-0001-6176-1952"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Chaehee","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136122459","display_name":"Minseok Seo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Seo, Minseok","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136164856","display_name":"Yeeun Seong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Seong, Yeeun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5058744764","display_name":"Doyi Kim","orcid":"https://orcid.org/0000-0002-9849-3127"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Doyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5069759184","display_name":"Changick Kim","orcid":"https://orcid.org/0000-0001-9323-8488"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Changick","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/T11948","display_name":"Machine Learning in Materials Science","score":0.26570001244544983,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.26570001244544983,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.23569999635219574,"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/T13910","display_name":"Computational and Text Analysis Methods","score":0.09910000115633011,"subfield":{"id":"https://openalex.org/subfields/3300","display_name":"General Social Sciences"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.717199981212616},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.5758000016212463},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.5476999878883362},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.546500027179718},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.44020000100135803},{"id":"https://openalex.org/keywords/data-modeling","display_name":"Data modeling","score":0.33809998631477356}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8105000257492065},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.717199981212616},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5758000016212463},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.5476999878883362},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.546500027179718},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5065000057220459},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4406000077724457},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.44020000100135803},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.33809998631477356},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.32919999957084656},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.2969000041484833},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.29490000009536743},{"id":"https://openalex.org/C179603123","wikidata":"https://www.wikidata.org/wiki/Q1941921","display_name":"Modeling language","level":3,"score":0.2732999920845032},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.26980000734329224},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.2529999911785126}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.13369","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.13369","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.2605.13369","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.13369","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/4","display_name":"Quality Education","score":0.573095440864563}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"are":[4],"typically":[5],"deployed":[6],"with":[7],"fixed":[8],"parameters,":[9],"and":[10,127,161,183],"their":[11],"performance":[12],"is":[13,103,141,180,192],"often":[14],"improved":[15],"by":[16,52],"allocating":[17],"more":[18],"computation":[19],"at":[20,135,194],"inference":[21,91],"time.":[22,137],"While":[23],"such":[24,124],"test-time":[25,171,187],"scaling":[26],"can":[27],"be":[28],"effective,":[29],"it":[30],"cannot":[31],"correct":[32],"model":[33,38,88,140],"misconceptions":[34],"or":[35,66],"adapt":[36],"the":[37,40,97,105,146,162],"to":[39,144],"specific":[41],"structure":[42],"of":[43],"an":[44,181],"individual":[45],"query.":[46,99],"Test-time":[47],"optimization":[48,172],"addresses":[49],"this":[50,76],"limitation":[51],"enabling":[53,149],"parameter":[54],"updates":[55],"during":[56,90],"inference,":[57],"but":[58],"existing":[59],"approaches":[60],"either":[61],"rely":[62],"on":[63,120],"external":[64,154],"data":[65],"optimize":[67],"generic":[68],"self-supervised":[69],"objectives":[70],"that":[71,86,104,177],"lack":[72],"query-specific":[73,150],"alignment.":[74],"In":[75],"work,":[77],"we":[78],"propose":[79],"Query-Conditioned":[80],"Test-Time":[81],"Self-Training":[82],"(QueST),":[83],"a":[84],"framework":[85],"adapts":[87],"parameters":[89],"using":[92],"supervision":[93,131],"derived":[94],"directly":[95],"from":[96],"input":[98,106],"Our":[100],"key":[101],"insight":[102],"query":[107],"itself":[108],"encodes":[109],"latent":[110],"signals":[111],"sufficient":[112],"for":[113,132,186],"constructing":[114],"structurally":[115],"related":[116],"problem--solution":[117],"pairs.":[118],"Based":[119],"this,":[121],"QueST":[122,167],"generates":[123],"query-conditioned":[125,178],"pairs":[126],"uses":[128],"them":[129],"as":[130],"parameter-efficient":[133],"fine-tuning":[134],"test":[136],"The":[138],"adapted":[139],"then":[142],"used":[143],"produce":[145],"final":[147],"answer,":[148],"adaptation":[151,188],"without":[152],"any":[153],"data.":[155],"Across":[156],"seven":[157],"mathematical":[158],"reasoning":[159,165],"benchmarks":[160],"GPQA-Diamond":[163],"scientific":[164],"benchmark,":[166],"consistently":[168],"outperforms":[169],"strong":[170],"baselines.":[173],"These":[174],"results":[175],"demonstrate":[176],"self-training":[179],"effective":[182],"practical":[184],"paradigm":[185],"in":[189],"LLMs.":[190],"Code":[191],"available":[193],"https://chssong.github.io/Query-Conditioned-TTST/.":[195]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-15T00:00:00"}