{"id":"https://openalex.org/W7161554750","doi":"https://doi.org/10.48550/arxiv.2605.15412","title":"From Feedback Loops to Policy Updates: Reinforcement Fine-Tuning for LLM-Based Alpha Factor Discovery","display_name":"From Feedback Loops to Policy Updates: Reinforcement Fine-Tuning for LLM-Based Alpha Factor Discovery","publication_year":2026,"publication_date":"2026-05-14","ids":{"openalex":"https://openalex.org/W7161554750","doi":"https://doi.org/10.48550/arxiv.2605.15412"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.15412","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.15412","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.2605.15412","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5136412282","display_name":"Lingzhe Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Lingzhe","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069358025","display_name":"Tong Jia","orcid":"https://orcid.org/0000-0002-5946-9829"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jia, Tong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045281526","display_name":"Yunpeng Zhai","orcid":"https://orcid.org/0000-0003-2246-8045"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhai, Yunpeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136386701","display_name":"Zixuan Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Zixuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136400168","display_name":"Chiming Duan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Duan, Chiming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102113773","display_name":"Minghua He","orcid":"https://orcid.org/0000-0003-4439-9810"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Minghua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136437713","display_name":"Philip S. Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Philip S.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136429458","display_name":"Ying Li","orcid":"https://orcid.org/0009-0004-5197-4152"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Ying","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/T11326","display_name":"Stock Market Forecasting Methods","score":0.6890000104904175,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T11326","display_name":"Stock Market Forecasting Methods","score":0.6890000104904175,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.043800000101327896,"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/T11674","display_name":"Sports Analytics and Performance","score":0.019899999722838402,"subfield":{"id":"https://openalex.org/subfields/2002","display_name":"Economics and Econometrics"},"field":{"id":"https://openalex.org/fields/20","display_name":"Economics, Econometrics and Finance"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/executable","display_name":"Executable","score":0.8420000076293945},{"id":"https://openalex.org/keywords/factor","display_name":"Factor (programming language)","score":0.6625000238418579},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.5667999982833862},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.5569000244140625},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.46799999475479126},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.4138000011444092}],"concepts":[{"id":"https://openalex.org/C160145156","wikidata":"https://www.wikidata.org/wiki/Q778586","display_name":"Executable","level":2,"score":0.8420000076293945},{"id":"https://openalex.org/C2781039887","wikidata":"https://www.wikidata.org/wiki/Q1391724","display_name":"Factor (programming language)","level":2,"score":0.6625000238418579},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6351000070571899},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.5667999982833862},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.5569000244140625},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.46799999475479126},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4138000011444092},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41359999775886536},{"id":"https://openalex.org/C186886427","wikidata":"https://www.wikidata.org/wiki/Q5441213","display_name":"Feedback loop","level":2,"score":0.40529999136924744},{"id":"https://openalex.org/C64943373","wikidata":"https://www.wikidata.org/wiki/Q2651003","display_name":"Alpha (finance)","level":4,"score":0.4050000011920929},{"id":"https://openalex.org/C201374245","wikidata":"https://www.wikidata.org/wiki/Q104534","display_name":"Digital subscriber line","level":2,"score":0.3952000141143799},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.38920000195503235},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.36329999566078186},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3337000012397766},{"id":"https://openalex.org/C2780898871","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Performance metric","level":2,"score":0.2847000062465668},{"id":"https://openalex.org/C41045048","wikidata":"https://www.wikidata.org/wiki/Q202843","display_name":"Linear programming","level":2,"score":0.25600001215934753}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.15412","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.15412","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.2605.15412","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.15412","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":{"Modern":[0],"quantitative":[1,130],"trading":[2],"increasingly":[3],"relies":[4],"on":[5,47,84,117,198],"systematic":[6],"models":[7],"to":[8,94,139],"extract":[9],"predictive":[10],"signals":[11],"from":[12],"large-scale":[13],"financial":[14],"data,":[15],"where":[16],"alpha":[17,112,221],"factor":[18,39,113,194,222,230],"discovery":[19,114,223],"plays":[20],"a":[21,110,136,184],"central":[22],"role":[23],"in":[24,37,124,183],"transforming":[25],"market":[26,201],"observations":[27],"into":[28,132],"tradable":[29],"signals.":[30],"Recent":[31],"LLM-based":[32,220],"methods":[33,81],"have":[34],"shown":[35],"promise":[36],"automating":[38],"generation,":[40],"but":[41],"most":[42],"of":[43,121,206,215],"them":[44,164],"still":[45],"rely":[46],"prompt-level":[48],"generation--evaluation--feedback":[49],"loops":[50],"for":[51],"iterative":[52],"optimization.":[53],"As":[54],"the":[55,125,170,191,204,211],"loop":[56],"becomes":[57],"longer,":[58],"repeatedly":[59],"appended":[60],"historical":[61,141],"candidates":[62],"and":[63,75,100,168,227],"feedback":[64,77,123],"can":[65],"cause":[66],"context":[67],"explosion,":[68],"increase":[69],"inference":[70],"cost,":[71],"dilute":[72],"useful":[73],"information,":[74],"introduce":[76],"drift.":[78],"Moreover,":[79],"these":[80,105],"often":[82],"depend":[83],"very":[85],"large":[86],"LLMs":[87],"whose":[88],"stable":[89],"generation":[90],"preferences":[91],"may":[92],"lead":[93],"structurally":[95],"similar":[96],"expressions,":[97,162],"redundant":[98],"candidates,":[99],"search":[101],"stagnation.":[102],"To":[103],"address":[104],"limitations,":[106],"we":[107],"propose":[108],"\\textsc{QuantEvolver},":[109,207],"self-evolving":[111],"framework":[115],"based":[116],"reinforcement":[118],"fine-tuning.":[119],"Instead":[120],"accumulating":[122],"prompt,":[126],"\\textsc{QuantEvolver}":[127,148],"converts":[128],"executable":[129,159],"evaluation":[131,213],"policy":[133],"updates,":[134],"enabling":[135],"Miner":[137,171],"LLM":[138,172],"internalize":[140],"optimization":[142],"experience":[143],"through":[144,165],"parameter":[145],"learning.":[146],"Specifically,":[147],"constructs":[149],"high-quality":[150,178],"seed":[151],"factors,":[152],"builds":[153],"diverse":[154],"seed--time-window":[155],"training":[156],"tasks,":[157],"generates":[158],"Factor":[160,186],"DSL":[161],"evaluates":[163],"Regime":[166],"Backtest,":[167],"optimizes":[169],"with":[173],"Diversity-Complementarity":[174],"Reward.":[175],"During":[176],"training,":[177],"factors":[179],"are":[180],"continuously":[181],"accumulated":[182],"Mined":[185],"Database,":[187],"which":[188,208],"serves":[189],"as":[190],"final":[192],"discovered":[193],"library.":[195],"Extensive":[196],"experiments":[197],"three":[199],"realistic":[200],"benchmarks":[202],"demonstrate":[203],"effectiveness":[205],"consistently":[209],"improves":[210],"primary":[212],"metric":[214],"each":[216],"task":[217],"over":[218],"existing":[219],"baselines,":[224],"produces":[225],"higher-quality":[226],"more":[228],"complementary":[229],"pools.":[231]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-19T00:00:00"}