{"id":"https://openalex.org/W7135223358","doi":"https://doi.org/10.48550/arxiv.2603.11679","title":"LLMs can construct powerful representations and streamline sample-efficient supervised learning","display_name":"LLMs can construct powerful representations and streamline sample-efficient supervised learning","publication_year":2026,"publication_date":"2026-03-12","ids":{"openalex":"https://openalex.org/W7135223358","doi":"https://doi.org/10.48550/arxiv.2603.11679"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.11679","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.11679","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.2603.11679","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5070408960","display_name":"\u0130lker Demirel","orcid":"https://orcid.org/0000-0003-1035-8500"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Demirel, Ilker","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111482315","display_name":"Larry Shi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shi, Lawrence","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032905393","display_name":"Zeshan Hussain","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hussain, Zeshan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5013431623","display_name":"David Sontag","orcid":"https://orcid.org/0000-0002-5034-7796"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sontag, 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/T13702","display_name":"Machine Learning in Healthcare","score":0.9243999719619751,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.9243999719619751,"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":0.027000000700354576,"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/T11710","display_name":"Biomedical Text Mining and Ontologies","score":0.004699999932199717,"subfield":{"id":"https://openalex.org/subfields/1312","display_name":"Molecular Biology"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/rubric","display_name":"Rubric","score":0.906000018119812},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.7581999897956848},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.7566999793052673},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.6033999919891357},{"id":"https://openalex.org/keywords/downstream","display_name":"Downstream (manufacturing)","score":0.5364999771118164},{"id":"https://openalex.org/keywords/supervised-learning","display_name":"Supervised learning","score":0.477400004863739},{"id":"https://openalex.org/keywords/feature-learning","display_name":"Feature learning","score":0.46540001034736633},{"id":"https://openalex.org/keywords/labeled-data","display_name":"Labeled data","score":0.39169999957084656}],"concepts":[{"id":"https://openalex.org/C111640148","wikidata":"https://www.wikidata.org/wiki/Q847349","display_name":"Rubric","level":2,"score":0.906000018119812},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7698000073432922},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.7581999897956848},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.7566999793052673},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6439999938011169},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.6033999919891357},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5543000102043152},{"id":"https://openalex.org/C2776207758","wikidata":"https://www.wikidata.org/wiki/Q5303302","display_name":"Downstream (manufacturing)","level":2,"score":0.5364999771118164},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.477400004863739},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.46540001034736633},{"id":"https://openalex.org/C2776145971","wikidata":"https://www.wikidata.org/wiki/Q30673951","display_name":"Labeled data","level":2,"score":0.39169999957084656},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.3395000100135803},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.32589998841285706},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2980000078678131},{"id":"https://openalex.org/C2780966255","wikidata":"https://www.wikidata.org/wiki/Q5474306","display_name":"Foundation (evidence)","level":2,"score":0.2897000014781952},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.28439998626708984},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.27469998598098755},{"id":"https://openalex.org/C58973888","wikidata":"https://www.wikidata.org/wiki/Q1041418","display_name":"Semi-supervised learning","level":2,"score":0.27300000190734863},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2554999887943268},{"id":"https://openalex.org/C116409475","wikidata":"https://www.wikidata.org/wiki/Q1385056","display_name":"External Data Representation","level":2,"score":0.25440001487731934}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.11679","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.11679","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.2603.11679","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.11679","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":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"As":[0],"real-world":[1],"datasets":[2],"become":[3],"more":[4,85,133],"complex":[5],"and":[6,25,69,123,150],"heterogeneous,":[7],"supervised":[8],"learning":[9],"is":[10,74],"often":[11,28],"bottlenecked":[12],"by":[13,102],"input":[14,53],"representation":[15],"design.":[16],"Modeling":[17],"multimodal":[18],"data,":[19],"such":[20,141],"as":[21,63,142],"time-series,":[22],"free":[23],"text,":[24],"structured":[26],"records,":[27],"requires":[29],"non-trivial":[30],"domain":[31],"expertise.":[32],"We":[33,91],"propose":[34],"an":[35,43,103],"agentic":[36],"pipeline":[37],"to":[38,56,77,145],"streamline":[39],"this":[40],"process.":[41],"First,":[42],"LLM":[44,121],"analyzes":[45],"a":[46,58,64,84,124],"small":[47],"but":[48],"diverse":[49],"subset":[50],"of":[51,81,131],"text-serialized":[52],"examples":[54],"in-context":[55],"synthesize":[57],"global":[59],"rubric,":[60],"which":[61,96],"acts":[62],"programmatic":[65],"specification":[66],"for":[67,88],"extracting":[68],"organizing":[70],"evidence.":[71],"This":[72],"rubric":[73,114],"then":[75],"used":[76],"transform":[78],"naive":[79,120],"text-serializations":[80],"inputs":[82],"into":[83],"standardized":[86],"format":[87],"downstream":[89],"models.":[90],"also":[92],"describe":[93],"local":[94],"rubrics,":[95],"are":[97],"task-conditioned":[98],"interpretive":[99],"summaries":[100],"generated":[101],"LLM.":[104],"Across":[105],"15":[106],"clinical":[107,125],"tasks":[108],"from":[109],"the":[110],"EHRSHOT":[111],"benchmark,":[112],"our":[113],"approaches":[115],"significantly":[116],"outperform":[117],"count-feature":[118],"models,":[119],"baselines,":[122],"foundation":[126],"model":[127],"pretrained":[128],"on":[129],"orders":[130],"magnitude":[132],"data.":[134],"Beyond":[135],"performance,":[136],"rubrics":[137],"offer":[138],"operational":[139],"advantages":[140],"being":[143],"easy":[144],"audit,":[146],"cost-effectiveness":[147],"at":[148],"scale,":[149],"facilitating":[151],"tabular":[152],"representations.":[153]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-14T00:00:00"}