{"id":"https://openalex.org/W7161015768","doi":"https://doi.org/10.48550/arxiv.2605.11518","title":"AutoLLMResearch: Training Research Agents for Automating LLM Experiment Configuration - Learning from Cheap, Optimizing Expensive","display_name":"AutoLLMResearch: Training Research Agents for Automating LLM Experiment Configuration - Learning from Cheap, Optimizing Expensive","publication_year":2026,"publication_date":"2026-05-12","ids":{"openalex":"https://openalex.org/W7161015768","doi":"https://doi.org/10.48550/arxiv.2605.11518"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.11518","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.11518","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":null,"license_id":null,"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.2605.11518","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5091542874","display_name":"Taicheng Guo","orcid":"https://orcid.org/0000-0001-7919-6912"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Taicheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136025506","display_name":"Nitesh V. Chawla","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chawla, Nitesh V.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136038567","display_name":"Olaf Wiest","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wiest, Olaf","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136087251","display_name":"Xiangliang Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Xiangliang","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/T11948","display_name":"Machine Learning in Materials Science","score":0.23440000414848328,"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":0.23440000414848328,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.17589999735355377,"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":0.055399999022483826,"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/scalability","display_name":"Scalability","score":0.6844000220298767},{"id":"https://openalex.org/keywords/interpretability","display_name":"Interpretability","score":0.5870000123977661},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.5327000021934509},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5300999879837036},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.47870001196861267},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.46070000529289246},{"id":"https://openalex.org/keywords/reuse","display_name":"Reuse","score":0.4153999984264374},{"id":"https://openalex.org/keywords/markov-decision-process","display_name":"Markov decision process","score":0.4088999927043915},{"id":"https://openalex.org/keywords/workflow","display_name":"Workflow","score":0.3935000002384186}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.795799970626831},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6844000220298767},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.5870000123977661},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5852000117301941},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.5327000021934509},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5300999879837036},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.47870001196861267},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.46070000529289246},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.42320001125335693},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.4153999984264374},{"id":"https://openalex.org/C106189395","wikidata":"https://www.wikidata.org/wiki/Q176789","display_name":"Markov decision process","level":3,"score":0.4088999927043915},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.3935000002384186},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.3865000009536743},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.3635999858379364},{"id":"https://openalex.org/C34559072","wikidata":"https://www.wikidata.org/wiki/Q2334061","display_name":"Design of experiments","level":2,"score":0.36010000109672546},{"id":"https://openalex.org/C85847156","wikidata":"https://www.wikidata.org/wiki/Q59015987","display_name":"Verifiable secret sharing","level":3,"score":0.3424000144004822},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.3264000117778778},{"id":"https://openalex.org/C62230096","wikidata":"https://www.wikidata.org/wiki/Q275969","display_name":"Crowdsourcing","level":2,"score":0.3253999948501587},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.29249998927116394},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.28630000352859497},{"id":"https://openalex.org/C2778049539","wikidata":"https://www.wikidata.org/wiki/Q17002908","display_name":"Bayesian optimization","level":2,"score":0.2854999899864197},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.28380000591278076},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.2808000147342682},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.2777000069618225},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.2761000096797943},{"id":"https://openalex.org/C132459708","wikidata":"https://www.wikidata.org/wiki/Q744069","display_name":"Extrapolation","level":2,"score":0.275299996137619},{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.26440000534057617},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.26339998841285706},{"id":"https://openalex.org/C2777362162","wikidata":"https://www.wikidata.org/wiki/Q5594431","display_name":"Grand Challenges","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C2777402240","wikidata":"https://www.wikidata.org/wiki/Q6783436","display_name":"Masking (illustration)","level":2,"score":0.2572999894618988},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.25290000438690186},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.250900000333786}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.11518","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.11518","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.11518","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.11518","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":null,"license_id":null,"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":{"Effectively":[0],"configuring":[1],"scalable":[2,54,229],"large":[3],"language":[4],"model":[5],"(LLM)":[6],"experiments,":[7],"spanning":[8],"architecture":[9],"design,":[10],"hyperparameter":[11],"tuning,":[12],"and":[13,30,49,82,108,195,214,225],"beyond,":[14],"is":[15,51,122],"crucial":[16],"for":[17,43,60,228],"advancing":[18],"LLM":[19,55,75,117,143,166,231],"research,":[20],"as":[21,189,222],"poor":[22],"configuration":[23,144,187],"choices":[24],"can":[25],"waste":[26],"substantial":[27],"computational":[28],"resources":[29],"prevent":[31],"models":[32],"from":[33,105],"realizing":[34],"their":[35],"full":[36],"potential.":[37],"Prior":[38],"automated":[39],"methods":[40],"are":[41,57],"designed":[42],"low-cost":[44],"settings":[45],"where":[46],"repeated":[47],"trial":[48],"error":[50],"feasible,":[52],"but":[53],"experiments":[56,107,209],"too":[58],"expensive":[59,116],"such":[61],"extensive":[62],"iteration.":[63],"To":[64,146],"our":[65,217],"knowledge,":[66],"no":[67],"work":[68],"has":[69],"addressed":[70],"the":[71,139,142,211],"automation":[72],"of":[73,141,176,216],"high-cost":[74],"experiment":[76,167,178,232],"configurations,":[77],"leaving":[78],"this":[79,89],"problem":[80],"labor-intensive":[81],"dependent":[83],"on":[84,207],"expert":[85],"intuition.":[86],"Motivated":[87],"by":[88,170],"gap,":[90],"we":[91,149],"propose":[92,150],"AutoLLMResearch,":[93],"an":[94,126],"agentic":[95],"framework":[96,153],"that":[97,137,185],"mimics":[98],"how":[99,123],"human":[100],"researchers":[101],"learn":[102],"generalizable":[103],"principles":[104],"low-fidelity":[106],"extrapolate":[109],"to":[110,124,128],"efficiently":[111],"identify":[112],"promising":[113],"configurations":[114],"in":[115],"settings.":[118],"The":[119],"core":[120],"challenge":[121],"enable":[125],"agent":[127],"learn,":[129],"through":[130],"interaction":[131],"with":[132,154],"a":[133,151,160,190,223],"multi-fidelity":[134,161],"experimental":[135],"environment":[136,162],"captures":[138],"structure":[140],"landscape.":[145],"achieve":[147],"this,":[148],"systematic":[152],"two":[155],"key":[156],"components:":[157],"1)":[158],"LLMConfig-Gym,":[159],"encompassing":[163],"four":[164],"critical":[165],"tasks,":[168],"supported":[169],"over":[171],"one":[172],"million":[173],"GPU":[174],"hours":[175],"verifiable":[177],"outcomes;":[179],"2)":[180],"A":[181],"structured":[182],"training":[183],"pipeline":[184],"formulates":[186],"research":[188],"long-horizon":[191],"Markov":[192],"Decision":[193],"Process":[194],"accordingly":[196],"incentivizes":[197],"cross-fidelity":[198],"extrapolation":[199],"reasoning.":[200],"Extensive":[201],"evaluation":[202],"against":[203],"diverse":[204],"strong":[205],"baselines":[206],"held-out":[208],"demonstrates":[210],"effectiveness,":[212],"generalization,":[213],"interpretability":[215],"framework,":[218],"supporting":[219],"its":[220],"potential":[221],"practical":[224],"general":[226],"solution":[227],"real-world":[230],"automation.":[233]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-14T00:00:00"}