{"id":"https://openalex.org/W7162646326","doi":"https://doi.org/10.48550/arxiv.2605.28057","title":"On the Learnability of Test-Time Adaptation: A Recovery Complexity Perspective","display_name":"On the Learnability of Test-Time Adaptation: A Recovery Complexity Perspective","publication_year":2026,"publication_date":"2026-05-27","ids":{"openalex":"https://openalex.org/W7162646326","doi":"https://doi.org/10.48550/arxiv.2605.28057"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.28057","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.28057","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2605.28057","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137216395","display_name":"Zhi Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Zhi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137232662","display_name":"Ming Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Ming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137197977","display_name":"Shi-Yu Tian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tian, Shi-Yu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137223858","display_name":"Kun-Yang Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Kun-Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137237313","display_name":"Lan-Zhe Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Lan-Zhe","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137262217","display_name":"Yu-Feng Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yu-Feng","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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.5828999876976013,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.5828999876976013,"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/T12761","display_name":"Data Stream Mining Techniques","score":0.2401999980211258,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.035999998450279236,"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/learnability","display_name":"Learnability","score":0.9483000040054321},{"id":"https://openalex.org/keywords/complement","display_name":"Complement (music)","score":0.6474999785423279},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.573199987411499},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.5597000122070312},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.5170000195503235},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.43209999799728394}],"concepts":[{"id":"https://openalex.org/C2777723229","wikidata":"https://www.wikidata.org/wiki/Q4367921","display_name":"Learnability","level":2,"score":0.9483000040054321},{"id":"https://openalex.org/C112313634","wikidata":"https://www.wikidata.org/wiki/Q7886648","display_name":"Complement (music)","level":5,"score":0.6474999785423279},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5752000212669373},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.573199987411499},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.5597000122070312},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.5170000195503235},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.4668999910354614},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.43209999799728394},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.3971000015735626},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.38609999418258667},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3531000018119812},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3305000066757202},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.2953999936580658},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.287200003862381},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.2524999976158142}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.28057","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.28057","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2605.28057","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.28057","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Test-time":[0],"adaptation":[1],"(TTA)":[2],"aims":[3],"to":[4,7,91,106],"adapt":[5],"models":[6],"maintain":[8,92],"reliable":[9],"performance":[10],"on":[11,149],"non-stationary":[12,28,126],"test":[13,127],"streams":[14,29],"without":[15],"requiring":[16],"labeled":[17],"data.":[18],"Despite":[19],"its":[20],"empirical":[21],"success,":[22],"the":[23,36,47,67,73,87,111],"learnability":[24,74,166],"of":[25,38,75,114,135,155],"TTA":[26,48,76,107,156,169],"under":[27],"remains":[30],"unexplored.":[31],"A":[32],"key":[33],"challenge":[34],"is":[35,103],"lack":[37],"a":[39,96,121,130],"principled":[40],"theoretical":[41,69],"framework":[42,70],"that":[43,170],"simultaneously":[44],"aligns":[45],"with":[46,99],"objective":[49],"and":[50,57,77,81,102,132,138,146,157],"captures":[51],"both":[52,136],"continuously":[53],"evolving":[54],"distribution":[55],"shifts":[56],"intrinsic":[58,159],"information":[59],"constraints.":[60],"To":[61],"address":[62],"this":[63,117],"gap,":[64],"we":[65,119],"propose":[66],"first":[68],"for":[71,125,168],"studying":[72],"introduce":[78,120],"$(\u03b5,\u03b4)$-Recovery":[79],"Complexity":[80],"$(\u03b5,\u03c1)$-TTA":[82],"Learnability.":[83],"Recovery":[84],"complexity":[85],"measures":[86,110],"post-shift":[88],"time":[89],"needed":[90],"excess":[93],"risk":[94],"below":[95],"target":[97],"level":[98],"high":[100],"probability,":[101],"further":[104],"extended":[105],"learnability,":[108],"which":[109],"long-term":[112],"reliability":[113],"TTA.":[115],"Within":[116],"framework,":[118],"novel":[122],"discrete":[123],"surrogate":[124],"streams,":[128],"enabling":[129],"unified":[131,165],"tractable":[133],"analysis":[134],"gradual":[137],"abrupt":[139],"shifts.":[140],"We":[141],"derive":[142],"order-wise":[143],"matching":[144],"lower":[145],"upper":[147],"bounds":[148],"recovery":[150],"complexity,":[151],"revealing":[152],"fundamental":[153],"limits":[154],"an":[158],"adaptivity-information":[160],"trade-off.":[161],"These":[162],"results":[163],"provide":[164],"guarantees":[167],"complement":[171],"regret-based":[172],"analyses.":[173]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-29T00:00:00"}