{"id":"https://openalex.org/W7203746987","doi":"https://doi.org/10.48550/arxiv.2608.17528","title":"Agent Lightning v1.0: Towards Harnessed Agentic RL","display_name":"Agent Lightning v1.0: Towards Harnessed Agentic RL","publication_year":2026,"publication_date":"2026-08-18","ids":{"openalex":"https://openalex.org/W7203746987","doi":"https://doi.org/10.48550/arxiv.2608.17528"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.17528","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.17528","pdf_url":"https://arxiv.org/pdf/2608.17528","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/2608.17528","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134809632","display_name":"Zhiyuan He","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Zhiyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zhang, Siwei","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Siwei","raw_affiliation_strings":["Fudan University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Fudan University","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147530115","display_name":"Zhiwen Zhou","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhou, Zhiwen","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127440551","display_name":"Yuqing Yang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Yuqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147510096","display_name":"Yu Kang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kang, Yu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147545653","display_name":"Yuge Zhang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Yuge","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124048587","display_name":"Luna K. Qiu","orcid":"https://orcid.org/0009-0000-6796-4494","observed_orcids":["https://orcid.org/0009-0000-6796-4494"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qiu, Luna K.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147448430","display_name":"Tin Yan Tsui","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tsui, Tin Yan","raw_affiliation_strings":["University of Edinburgh API Gateway"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Edinburgh API Gateway","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147614341","display_name":"Jiahang Xu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Jiahang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137615311","display_name":"Chong Luo","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Chong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.7285000085830688,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.7285000085830688,"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/T13904","display_name":"Artificial Intelligence Applications","score":0.03720000013709068,"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/T10260","display_name":"Software Engineering Research","score":0.03440000116825104,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/agent-harnesses","display_name":"agent harnesses","score":0.8849999904632568},{"id":"https://openalex.org/keywords/llm-based-agents","display_name":"LLM agents","score":0.5429999828338623},{"id":"https://openalex.org/keywords/ai-coding-agents","display_name":"coding agents","score":0.515999972820282},{"id":"https://openalex.org/keywords/swe-bench","display_name":"SWE-bench","score":0.6729999780654907},{"id":"https://openalex.org/keywords/search-agents","display_name":"search agents","score":0.9409999847412109},{"id":"https://openalex.org/keywords/training-stability","display_name":"training stability","score":0.49399998784065247}],"concepts":[{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.6031000018119812},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5612000226974487},{"id":"https://openalex.org/C31395832","wikidata":"https://www.wikidata.org/wiki/Q1318674","display_name":"Testbed","level":2,"score":0.4641999900341034},{"id":"https://openalex.org/C2780463512","wikidata":"https://www.wikidata.org/wiki/Q15122700","display_name":"Trainer","level":2,"score":0.42640000581741333},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.40529999136924744},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.38999998569488525},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.3472000062465668},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3467000126838684},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.3458999991416931},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.29739999771118164},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.29580000042915344},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.2800999879837036},{"id":"https://openalex.org/C74072328","wikidata":"https://www.wikidata.org/wiki/Q1142726","display_name":"Intelligent agent","level":2,"score":0.272599995136261},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.2700999975204468},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.26499998569488525},{"id":"https://openalex.org/C13687954","wikidata":"https://www.wikidata.org/wiki/Q4826847","display_name":"Autonomous agent","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C61423126","wikidata":"https://www.wikidata.org/wiki/Q187432","display_name":"Scripting language","level":2,"score":0.2599000036716461},{"id":"https://openalex.org/C2776544517","wikidata":"https://www.wikidata.org/wiki/Q189447","display_name":"Unexpected events","level":2,"score":0.25049999356269836}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.17528","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.17528","pdf_url":"https://arxiv.org/pdf/2608.17528","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.2608.17528","doi":"https://doi.org/10.48550/arxiv.2608.17528","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.17528","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:2608.17528","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.17528","pdf_url":"https://arxiv.org/pdf/2608.17528","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":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7203746987.pdf","grobid_xml":"https://content.openalex.org/works/W7203746987.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7077067785","https://openalex.org/W7170285085","https://openalex.org/W7204941889","https://openalex.org/W7162404462","https://openalex.org/W7160120919","https://openalex.org/W7203698138","https://openalex.org/W7106156359","https://openalex.org/W7223205430","https://openalex.org/W7155108762","https://openalex.org/W7154315933"],"abstract_inverted_index":{"Modern":[0],"agents":[1,34],"operate":[2],"inside":[3],"agent":[4,21,151],"harnesses":[5,152],"that":[6,31],"manage":[7],"tools,":[8],"context,":[9],"and":[10,56,118,127,153,169,172,186,208],"control":[11],"flow,":[12],"making":[13],"the":[14,20,68,85,89,93,98,205],"harness":[15,70],"a":[16,28,134,156,174,199],"critical":[17],"part":[18],"of":[19,103,146],"system.":[22],"Our":[23],"original":[24],"Agent":[25,131],"Lightning":[26,132],"introduced":[27],"disaggregated":[29],"architecture":[30],"connects":[32],"arbitrary":[33,150],"to":[35,60,197,211],"RL":[36,78,140,189],"training":[37,90,125,184,209],"through":[38],"an":[39,43],"LLM":[40,104],"endpoint":[41],"proxy,":[42],"approach":[44],"later":[45],"adopted":[46],"by":[47],"frameworks":[48],"such":[49],"as":[50,63,155],"verl":[51],"Uni-Agent,":[52],"AReaL":[53],"2.0,":[54],"slime,":[55],"Polar.":[57],"We":[58,129,163,203],"refer":[59],"this":[61],"paradigm":[62],"harnessed":[64,138,216],"agentic":[65,77,83,139,217],"RL,":[66],"where":[67],"deploy-time":[69],"directly":[71],"participates":[72],"in":[73,110,142],"model":[74],"post-training.":[75],"Harnessed":[76],"differs":[79],"fundamentally":[80],"from":[81,195],"traditional":[82],"RL:":[84],"harness,":[86],"rather":[87],"than":[88],"engine,":[91],"owns":[92],"environment":[94],"interaction":[95],"loop,":[96],"while":[97],"trainer":[99],"observes":[100],"only":[101,182],"sequences":[102],"request-response":[105],"pairs.":[106],"This":[107],"introduces":[108],"challenges":[109],"retokenization,":[111],"sample":[112],"merging,":[113],"advantage":[114],"calculation,":[115],"loss":[116],"normalization,":[117],"backend":[119],"scheduling,":[120],"which":[121],"can":[122],"substantially":[123],"affect":[124],"stability":[126],"effectiveness.":[128],"present":[130],"v1.0,":[133],"lightweight":[135],"framework":[136],"for":[137,159,178],"implemented":[141],"approximately":[143],"3,500":[144],"lines":[145],"code.":[147],"It":[148],"supports":[149],"serves":[154],"practical":[157],"testbed":[158],"studying":[160],"these":[161],"challenges.":[162],"evaluate":[164],"it":[165],"on":[166,192,215],"instruction-following,":[167],"search,":[168],"coding":[170],"agents,":[171],"provide":[173],"complete":[175,206],"reproducible":[176,213],"pipeline":[177],"coding-agent":[179],"RL.":[180,218],"Using":[181],"6K":[183],"examples":[185],"modest":[187],"compute,":[188],"improves":[190],"Qwen3.5-9B":[191],"SWE-bench":[193],"Verified":[194],"41.8%":[196],"56.4%,":[198],"14.6-point":[200],"absolute":[201],"gain.":[202],"release":[204],"workflow":[207],"scripts":[210],"facilitate":[212],"research":[214]},"counts_by_year":[],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2026-08-20T00:00:00"}