{"id":"https://openalex.org/W7160624798","doi":"https://doi.org/10.48550/arxiv.2605.05638","title":"Scaling Pretrained Representations Enables Label-Free Out-of-Distribution Detection Without Fine-Tuning","display_name":"Scaling Pretrained Representations Enables Label-Free Out-of-Distribution Detection Without Fine-Tuning","publication_year":2026,"publication_date":"2026-05-07","ids":{"openalex":"https://openalex.org/W7160624798","doi":"https://doi.org/10.48550/arxiv.2605.05638"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.05638","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.05638","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.2605.05638","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135707336","display_name":"Brett Barkley","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Barkley, Brett","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006569343","display_name":"Preston Culbertson","orcid":"https://orcid.org/0000-0002-1403-8697"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Culbertson, Preston","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135686094","display_name":"David Fridovich-Keil","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fridovich-Keil, David","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/T10036","display_name":"Advanced Neural Network Applications","score":0.24959999322891235,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.24959999322891235,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.218299999833107,"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":0.2087000012397766,"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/representation","display_name":"Representation (politics)","score":0.652899980545044},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.6191999912261963},{"id":"https://openalex.org/keywords/estimator","display_name":"Estimator","score":0.6043000221252441},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.5468000173568726},{"id":"https://openalex.org/keywords/detector","display_name":"Detector","score":0.5408999919891357},{"id":"https://openalex.org/keywords/mahalanobis-distance","display_name":"Mahalanobis distance","score":0.49410000443458557},{"id":"https://openalex.org/keywords/encode","display_name":"ENCODE","score":0.4237000048160553},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4056999981403351}],"concepts":[{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6933000087738037},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.652899980545044},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6439999938011169},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.6191999912261963},{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.6043000221252441},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.5468000173568726},{"id":"https://openalex.org/C94915269","wikidata":"https://www.wikidata.org/wiki/Q1834857","display_name":"Detector","level":2,"score":0.5408999919891357},{"id":"https://openalex.org/C1921717","wikidata":"https://www.wikidata.org/wiki/Q1334846","display_name":"Mahalanobis distance","level":2,"score":0.49410000443458557},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.4237000048160553},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4056999981403351},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3921999931335449},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.35019999742507935},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.34950000047683716},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.3379000127315521},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.3328000009059906},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.321399986743927},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.320499986410141},{"id":"https://openalex.org/C2776151529","wikidata":"https://www.wikidata.org/wiki/Q3045304","display_name":"Object detection","level":3,"score":0.3095000088214874},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.30410000681877136},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2955999970436096},{"id":"https://openalex.org/C137270730","wikidata":"https://www.wikidata.org/wiki/Q120811","display_name":"Detection theory","level":3,"score":0.29019999504089355},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2709999978542328},{"id":"https://openalex.org/C4679612","wikidata":"https://www.wikidata.org/wiki/Q866298","display_name":"Aggregate (composite)","level":2,"score":0.26420000195503235},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2547999918460846}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.05638","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.05638","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.2605.05638","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.05638","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Models":[0],"trained":[1],"with":[2,126],"deep":[3],"learning":[4],"often":[5,32],"fail":[6],"to":[7,18],"signal":[8],"when":[9],"inputs":[10],"fall":[11],"outside":[12],"their":[13,54,107],"training":[14],"data":[15],"manifold,":[16],"leading":[17],"unreliable":[19],"predictions":[20],"under":[21],"distribution":[22],"shift.":[23],"Prior":[24],"work":[25],"suggests":[26],"that":[27,53,150],"effective":[28],"out-of-distribution":[29],"(OOD)":[30],"detection":[31,109,153],"requires":[33],"class-conditional":[34],"modeling":[35],"or":[36],"specialized":[37],"models":[38,50],"obtained":[39],"through":[40],"supervised":[41],"fine-tuning.":[42],"We":[43],"revisit":[44],"this":[45],"assumption":[46],"in":[47],"modern":[48],"pretrained":[49,162],"and":[51,73,90,120,130,141,174],"show":[52],"frozen":[55,161,180],"representations":[56,145],"already":[57],"encode":[58],"sufficient":[59],"geometric":[60],"structure":[61],"for":[62],"accurate":[63],"label-free":[64,79,151],"OOD":[65,152],"detection.":[66],"Across":[67],"59":[68],"backbone-task":[69],"pairings":[70],"spanning":[71],"vision":[72,142],"language,":[74],"we":[75],"compare":[76],"two":[77,135],"complementary":[78],"detectors:":[80],"a":[81,92,103,114],"global":[82,121],"Mahalanobis":[83],"estimator":[84,96],"fit":[85],"on":[86,98,156,179],"unlabeled":[87],"latent":[88],"representations,":[89],"ReSCOPED,":[91],"lightweight,":[93],"diffusion-based":[94],"typicality":[95],"operating":[97],"the":[99,134,157,165],"same":[100],"features":[101],"at":[102],"local":[104,119],"level.":[105],"Despite":[106],"different":[108],"mechanisms,":[110],"representation":[111,128],"scaling":[112],"reveals":[113],"consistent":[115],"regime-dependent":[116],"pattern:":[117],"both":[118,139],"detectors'":[122],"absolute":[123],"performance":[124,131],"improves":[125],"better":[127],"quality,":[129],"gaps":[132],"between":[133],"detectors":[136],"disappear":[137],"across":[138],"language":[140],"tasks":[143],"as":[144,170],"scale.":[146],"These":[147],"results":[148],"suggest":[149],"depends":[154],"strongly":[155],"geometry":[158],"exposed":[159],"by":[160],"backbones,":[163],"reducing":[164],"importance":[166],"of":[167],"detector":[168],"choice":[169],"backbone":[171],"scale":[172],"increases":[173],"enabling":[175],"efficient":[176],"deployment":[177],"directly":[178],"models.":[181]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-09T00:00:00"}