{"id":"https://openalex.org/W4404392017","doi":"https://doi.org/10.48550/arxiv.2411.07279","title":"The Surprising Effectiveness of Test-Time Training for Few-Shot Learning","display_name":"The Surprising Effectiveness of Test-Time Training for Few-Shot Learning","publication_year":2024,"publication_date":"2024-11-11","ids":{"openalex":"https://openalex.org/W4404392017","doi":"https://doi.org/10.48550/arxiv.2411.07279"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2411.07279","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2411.07279","pdf_url":"https://arxiv.org/pdf/2411.07279","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2411.07279","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5071734030","display_name":"Ekin Aky\u00fcrek","orcid":"https://orcid.org/0000-0002-5166-4689","observed_orcids":["https://orcid.org/0000-0002-5166-4689"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Aky\u00fcrek, Ekin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040083642","display_name":"Mehul Damani","orcid":"https://orcid.org/0000-0001-5790-376X","observed_orcids":["https://orcid.org/0000-0001-5790-376X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Damani, Mehul","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zweiger, Adam","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zweiger, Adam","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5003617960","display_name":"Linlu Qiu","orcid":"https://orcid.org/0000-0002-1696-5419","observed_orcids":["https://orcid.org/0000-0002-1696-5419"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qiu, Linlu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Guo, Han","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Han","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Pari, Jyothish","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pari, Jyothish","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100693798","display_name":"Yoon Kim","orcid":"https://orcid.org/0000-0002-4837-8411","observed_orcids":["https://orcid.org/0000-0002-4837-8411"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Yoon","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5070829652","display_name":"Jacob Andreas","orcid":"https://orcid.org/0000-0002-3141-5845","observed_orcids":["https://orcid.org/0000-0002-3141-5845"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Andreas, Jacob","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":4,"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":0.9413999915122986,"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":0.9413999915122986,"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/T10028","display_name":"Topic Modeling","score":0.05090000107884407,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.0031999999191612005,"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/test-time-training","display_name":"test-time training","score":1.0},{"id":"https://openalex.org/keywords/few-shot-learning","display_name":"few-shot learning","score":0.9990000128746033},{"id":"https://openalex.org/keywords/language-models","display_name":"language models","score":0.7900000214576721},{"id":"https://openalex.org/keywords/in-context-learning","display_name":"in-context learning","score":0.949999988079071},{"id":"https://openalex.org/keywords/in-context-examples","display_name":"in-context examples","score":0.5630000233650208},{"id":"https://openalex.org/keywords/abstraction-and-reasoning-corpus","display_name":"Abstraction and Reasoning Corpus","score":0.8109999895095825},{"id":"https://openalex.org/keywords/program-synthesis","display_name":"program synthesis","score":0.8019999861717224},{"id":"https://openalex.org/keywords/reasoning","display_name":"reasoning","score":0.3869999945163727}],"concepts":[{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.6823176145553589},{"id":"https://openalex.org/C2777267654","wikidata":"https://www.wikidata.org/wiki/Q3519023","display_name":"Test (biology)","level":2,"score":0.6783050894737244},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.40642914175987244},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.39051875472068787},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3490179777145386},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.34103041887283325},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.11924749612808228},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.08680912852287292},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.07870328426361084},{"id":"https://openalex.org/C18903297","wikidata":"https://www.wikidata.org/wiki/Q7150","display_name":"Ecology","level":1,"score":0.0656178891658783}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2411.07279","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2411.07279","pdf_url":"https://arxiv.org/pdf/2411.07279","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2411.07279","doi":"https://doi.org/10.48550/arxiv.2411.07279","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2411.07279","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,"listed_in":[],"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":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2411.07279","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2411.07279","pdf_url":"https://arxiv.org/pdf/2411.07279","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7092291745","https://openalex.org/W7133232273","https://openalex.org/W7197025912","https://openalex.org/W7212385854","https://openalex.org/W7166841804","https://openalex.org/W4416591542","https://openalex.org/W7106692097","https://openalex.org/W7092213973","https://openalex.org/W7166890186","https://openalex.org/W6967130704"],"abstract_inverted_index":{"Language":[0],"models":[1],"(LMs)":[2],"have":[3],"shown":[4],"impressive":[5],"performance":[6],"on":[7,89,114],"tasks":[8,19,142],"within":[9],"their":[10],"training":[11,36,149],"distribution,":[12],"but":[13],"often":[14],"struggle":[15],"with":[16,73,94,102],"structurally":[17],"novel":[18,141],"even":[20],"when":[21,100],"given":[22],"a":[23,46,54],"small":[24],"number":[25],"of":[26,34,137,147],"in-context":[27,74,115,138],"task":[28],"examples.":[29],"We":[30],"investigate":[31],"the":[32,65,90,122,135,145],"effectiveness":[33],"test-time":[35,148],"(TTT)":[37],"--":[38,52,86],"temporarily":[39],"updating":[40],"model":[41,153],"parameters":[42],"during":[43],"inference":[44],"using":[45],"loss":[47],"derived":[48],"from":[49],"input":[50],"data":[51],"as":[53],"mechanism":[55],"for":[56,140],"improving":[57],"LMs'":[58],"reasoning":[59],"and":[60,67,98,143],"few-shot":[61,119],"learning":[62,139],"capabilities.":[63],"On":[64,109],"Abstraction":[66],"Reasoning":[68],"Corpus":[69],"(ARC),":[70],"performing":[71],"TTT":[72,113],"examples":[75,116],"yields":[76],"up":[77],"to":[78,83,130,150],"$6\\times$":[79],"higher":[80],"accuracy":[81],"compared":[82],"fine-tuned":[84],"baselines":[85],"reaching":[87],"$53.0\\%$":[88],"public":[91],"validation":[92],"set":[93],"an":[95],"8B-parameter":[96],"LM":[97],"$61.9\\%$":[99],"ensembled":[101],"program-synthesis":[103],"methods,":[104],"matching":[105],"average":[106],"human":[107],"performance.":[108],"BIG-Bench":[110],"Hard":[111],"(BBH),":[112],"surpasses":[117],"standard":[118],"prompting":[120],"in":[121],"$10$-shot":[123],"setting":[124],"by":[125],"$7.3$":[126],"percentage":[127],"points":[128],"($50.5\\%$":[129],"$57.8\\%$).":[131],"Our":[132],"findings":[133],"highlight":[134],"limitations":[136],"demonstrate":[144],"potential":[146],"enhance":[151],"language":[152],"adaptability.":[154]},"counts_by_year":[{"year":2025,"cited_by_count":4}],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2024-11-15T00:00:00"}