{"id":"https://openalex.org/W7154621186","doi":"https://doi.org/10.48550/arxiv.2604.14116","title":"TREX: Automating LLM Fine-tuning via Agent-Driven Tree-based Exploration","display_name":"TREX: Automating LLM Fine-tuning via Agent-Driven Tree-based Exploration","publication_year":2026,"publication_date":"2026-04-15","ids":{"openalex":"https://openalex.org/W7154621186","doi":"https://doi.org/10.48550/arxiv.2604.14116"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.14116","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.14116","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.14116","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133820040","display_name":"Zerun Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Ma, Zerun","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133789364","display_name":"Guoqiang Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Guoqiang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125744015","display_name":"Xinchen Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Xinchen","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133785939","display_name":"Yicheng Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Yicheng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101594097","display_name":"He Du","orcid":"https://orcid.org/0000-0002-9547-1459"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Du, He","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133757271","display_name":"Bowen Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Bowen","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133760717","display_name":"Yanan Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Yanan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130809208","display_name":"Wenran Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Wenran","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133804969","display_name":"Kai Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Kai","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133747018","display_name":"Yining Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yining","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":10,"corresponding_author_ids":["https://openalex.org/A5133820040"],"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":0.17560000717639923,"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":0.17560000717639923,"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":0.1386999934911728,"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/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.13030000030994415,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.6628999710083008},{"id":"https://openalex.org/keywords/reuse","display_name":"Reuse","score":0.6517000198364258},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.6187000274658203},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.5871999859809875},{"id":"https://openalex.org/keywords/plan","display_name":"Plan (archaeology)","score":0.5672000050544739},{"id":"https://openalex.org/keywords/iterative-and-incremental-development","display_name":"Iterative and incremental development","score":0.45089998841285706},{"id":"https://openalex.org/keywords/automation","display_name":"Automation","score":0.3828999996185303},{"id":"https://openalex.org/keywords/core","display_name":"Core (optical fiber)","score":0.3564000129699707}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6811000108718872},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.6628999710083008},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.6517000198364258},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.6187000274658203},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5871999859809875},{"id":"https://openalex.org/C2776505523","wikidata":"https://www.wikidata.org/wiki/Q4785468","display_name":"Plan (archaeology)","level":2,"score":0.5672000050544739},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.5414999723434448},{"id":"https://openalex.org/C143587482","wikidata":"https://www.wikidata.org/wiki/Q1543216","display_name":"Iterative and incremental development","level":2,"score":0.45089998841285706},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4278999865055084},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.3828999996185303},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.3564000129699707},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.33079999685287476},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.32659998536109924},{"id":"https://openalex.org/C2780366209","wikidata":"https://www.wikidata.org/wiki/Q5170200","display_name":"Core model","level":2,"score":0.3215999901294708},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3190000057220459},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3181000053882599},{"id":"https://openalex.org/C2776904630","wikidata":"https://www.wikidata.org/wiki/Q356336","display_name":"Adept","level":3,"score":0.31610000133514404},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.30550000071525574},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.2752000093460083},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.27379998564720154},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.27250000834465027},{"id":"https://openalex.org/C145644426","wikidata":"https://www.wikidata.org/wiki/Q169411","display_name":"Unified Modeling Language","level":3,"score":0.27230000495910645},{"id":"https://openalex.org/C133462117","wikidata":"https://www.wikidata.org/wiki/Q4929239","display_name":"Data collection","level":2,"score":0.2709999978542328},{"id":"https://openalex.org/C107645828","wikidata":"https://www.wikidata.org/wiki/Q12070446","display_name":"System model","level":2,"score":0.2596000134944916},{"id":"https://openalex.org/C48262172","wikidata":"https://www.wikidata.org/wiki/Q16908765","display_name":"Design process","level":3,"score":0.25270000100135803}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.14116","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.14116","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.14116","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.14116","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":"article"},"sustainable_development_goals":[{"score":0.525130569934845,"display_name":"Responsible consumption and production","id":"https://metadata.un.org/sdg/12"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"While":[0],"Large":[1],"Language":[2],"Models":[3],"(LLMs)":[4],"have":[5],"empowered":[6],"AI":[7],"research":[8],"agents":[9],"to":[10,90,131],"perform":[11],"isolated":[12],"scientific":[13],"tasks,":[14],"automating":[15],"complex,":[16],"real-world":[17,123],"workflows,":[18],"such":[19],"as":[20,83],"LLM":[21,40,111],"training,":[22,112],"remains":[23],"a":[24,33,84,116],"significant":[25],"challenge.":[26],"In":[27],"this":[28],"paper,":[29],"we":[30,113],"introduce":[31],"TREX,":[32],"multi-agent":[34],"system":[35,54,89],"that":[36,140],"automates":[37],"the":[38,52,88,107,141],"entire":[39],"training":[41,66,74],"life-cycle.":[42],"By":[43],"orchestrating":[44],"collaboration":[45],"between":[46],"two":[47],"core":[48],"modules-the":[49],"Researcher":[50],"and":[51,61,72,75,98],"Executor-the":[53],"seamlessly":[55],"performs":[56],"requirement":[57],"analysis,":[58],"open-domain":[59],"literature":[60],"data":[62,70],"research,":[63],"formulation":[64],"of":[65,69,109],"strategies,":[67],"preparation":[68],"recipes,":[71],"model":[73,129,146],"evaluation.":[76],"The":[77],"multi-round":[78],"experimental":[79],"process":[80],"is":[81],"modeled":[82],"search":[85],"tree,":[86],"enabling":[87],"efficiently":[91],"plan":[92],"exploration":[93],"paths,":[94],"reuse":[95],"historical":[96],"results,":[97],"distill":[99],"high-level":[100],"insights":[101],"from":[102,122,126],"iterative":[103],"trials.":[104],"To":[105],"evaluate":[106],"capability":[108],"automated":[110],"construct":[114],"FT-Bench,":[115],"benchmark":[117],"comprising":[118],"10":[119],"tasks":[120],"derived":[121],"scenarios,":[124],"ranging":[125],"optimizing":[127],"fundamental":[128],"capabilities":[130],"enhancing":[132],"performance":[133,147],"on":[134,148],"domain-specific":[135],"tasks.":[136,150],"Experimental":[137],"results":[138],"demonstrate":[139],"TREX":[142],"agent":[143],"consistently":[144],"optimizes":[145],"target":[149]},"counts_by_year":[],"updated_date":"2026-04-17T06:04:52.305304","created_date":"2026-04-17T00:00:00"}