{"id":"https://openalex.org/W7220436573","doi":"https://doi.org/10.48550/arxiv.2610.02616","title":"VERSE: Verified Self-Evolving Optimizer for Agent Harnesses","display_name":"VERSE: Verified Self-Evolving Optimizer for Agent Harnesses","publication_year":2026,"publication_date":"2026-10-02","ids":{"openalex":"https://openalex.org/W7220436573","doi":"https://doi.org/10.48550/arxiv.2610.02616"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.02616","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.02616","pdf_url":"https://arxiv.org/pdf/2610.02616","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2610.02616","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101602470","display_name":"Zekai Wang","orcid":"https://orcid.org/0009-0002-4354-4705","observed_orcids":["https://orcid.org/0009-0002-4354-4705","https://orcid.org/0000-0001-5874-7092"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zekai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020526977","display_name":"Yingqiang Ge","orcid":"https://orcid.org/0000-0002-3736-2377","observed_orcids":["https://orcid.org/0000-0002-3736-2377"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ge, Yingqiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5157642918","display_name":"Zekun Wang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zekun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5157375883","display_name":"Hai Wang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Hai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134514053","display_name":"Yuhui Xu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Yuhui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5157486624","display_name":"Joshua Frandsen","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Frandsen, Joshua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5157393927","display_name":"Shancong Fu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fu, Shancong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013415067","display_name":"Ashia Wilson","orcid":"https://orcid.org/0000-0001-7072-2912","observed_orcids":["https://orcid.org/0000-0001-7072-2912"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wilson, Ashia C.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5001022750","display_name":"Chandan K. A. Reddy","orcid":"https://orcid.org/0000-0003-2839-3662","observed_orcids":["https://orcid.org/0000-0003-2839-3662"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Reddy, Chandan K.","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":true,"cited_by_count":4,"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/T10260","display_name":"Software Engineering Research","score":0.7846999764442444,"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"}},"topics":[{"id":"https://openalex.org/T10260","display_name":"Software Engineering Research","score":0.7846999764442444,"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"}},{"id":"https://openalex.org/T11975","display_name":"Evolutionary Algorithms and Applications","score":0.07919999957084656,"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/T10743","display_name":"Software Testing and Debugging Techniques","score":0.05719999969005585,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"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/llm-based-agents","display_name":"LLM agents","score":0.6349999904632568},{"id":"https://openalex.org/keywords/agent-harnesses","display_name":"agent harnesses","score":0.7139999866485596},{"id":"https://openalex.org/keywords/failure-analysis","display_name":"failure analysis","score":0.7229999899864197},{"id":"https://openalex.org/keywords/prompt-optimization","display_name":"prompt optimization","score":0.4560000002384186},{"id":"https://openalex.org/keywords/code-generation","display_name":"code generation","score":0.34700000286102295},{"id":"https://openalex.org/keywords/multi-agent-systems","display_name":"multi-agent systems","score":0.5149999856948853}],"concepts":[{"id":"https://openalex.org/C180591056","wikidata":"https://www.wikidata.org/wiki/Q654437","display_name":"Executor","level":2,"score":0.886900007724762},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7437999844551086},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.6550999879837036},{"id":"https://openalex.org/C2780385302","wikidata":"https://www.wikidata.org/wiki/Q367158","display_name":"Protocol (science)","level":3,"score":0.454800009727478},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.38580000400543213},{"id":"https://openalex.org/C2777267654","wikidata":"https://www.wikidata.org/wiki/Q3519023","display_name":"Test (biology)","level":2,"score":0.3634999990463257},{"id":"https://openalex.org/C128942645","wikidata":"https://www.wikidata.org/wiki/Q1568346","display_name":"Test case","level":3,"score":0.34290000796318054},{"id":"https://openalex.org/C45340560","wikidata":"https://www.wikidata.org/wiki/Q215382","display_name":"Disjoint sets","level":2,"score":0.34200000762939453},{"id":"https://openalex.org/C151552104","wikidata":"https://www.wikidata.org/wiki/Q7705809","display_name":"Test suite","level":4,"score":0.298799991607666},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.28529998660087585},{"id":"https://openalex.org/C2775936607","wikidata":"https://www.wikidata.org/wiki/Q466845","display_name":"Tracking (education)","level":2,"score":0.2759999930858612}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.02616","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.02616","pdf_url":"https://arxiv.org/pdf/2610.02616","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.2610.02616","doi":"https://doi.org/10.48550/arxiv.2610.02616","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.02616","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:2610.02616","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.02616","pdf_url":"https://arxiv.org/pdf/2610.02616","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/W7220436573.pdf","grobid_xml":"https://content.openalex.org/works/W7220436573.grobid-xml"},"referenced_works_count":28,"referenced_works":["https://openalex.org/W7140086132","https://openalex.org/W7141388408","https://openalex.org/W7154201446","https://openalex.org/W7155562078","https://openalex.org/W7158100772","https://openalex.org/W7160849836","https://openalex.org/W7160908075","https://openalex.org/W7161113800","https://openalex.org/W7162554468","https://openalex.org/W7162556122","https://openalex.org/W7163428483","https://openalex.org/W7163526135","https://openalex.org/W7163637799","https://openalex.org/W7164008618","https://openalex.org/W7164143017","https://openalex.org/W7164848732","https://openalex.org/W7167605568","https://openalex.org/W7167622861","https://openalex.org/W7167803489","https://openalex.org/W7169776101","https://openalex.org/W7172337529","https://openalex.org/W7202208313","https://openalex.org/W7203920666","https://openalex.org/W7204215770","https://openalex.org/W7206206100","https://openalex.org/W7212048886","https://openalex.org/W7213359549","https://openalex.org/W7214002283"],"related_works":["https://openalex.org/W7172337529","https://openalex.org/W7216595909","https://openalex.org/W7202208313","https://openalex.org/W7164008618","https://openalex.org/W7212406378","https://openalex.org/W7201862772","https://openalex.org/W7202216241","https://openalex.org/W7202196053","https://openalex.org/W7204215770","https://openalex.org/W7131850018"],"abstract_inverted_index":{"Harness":[0],"evolution":[1],"improves":[2,170],"an":[3,23],"LLM":[4],"agent's":[5],"prompts,":[6,141],"tools,":[7,143],"and":[8,15,40,89,115,124,138,145,153,166,180,193,199],"workflow,":[9],"while":[10,121,147],"the":[11,64,108,131,135,148,151,202],"optimizer's":[12],"own":[13,81,140],"tools":[14,82],"procedures":[16],"often":[17],"remain":[18],"fixed.":[19,157],"We":[20],"study":[21,69],"whether":[22],"optimizer":[24,53,102,109,132,152],"can":[25],"improve":[26,57],"another":[27],"agent":[28,104],"more":[29],"effectively":[30],"by":[31,93],"also":[32],"improving":[33],"how":[34],"it":[35],"diagnoses":[36],"failures,":[37,114],"develops":[38],"edits,":[39,112],"tests":[41],"their":[42,80],"effects.":[43],"Two":[44],"observations":[45],"guide":[46],"our":[47],"design.":[48],"In":[49],"a":[50,99,159],"controlled":[51],"study,":[52],"self-evolution":[54],"fails":[55],"to":[56],"performance":[58],"without":[59],"execution-based":[60],"verification,":[61,86],"but":[62],"achieves":[63],"best":[65,188],"result":[66],"of":[67,150],"that":[68],"when":[70],"verification":[71],"is":[72,206],"available.":[73],"Across":[74],"five":[75,185],"executors,":[76],"self-evolving":[77],"optimizers":[78,175],"build":[79],"for":[83,103,201],"failure":[84],"analysis,":[85],"training":[87],"audits,":[88],"workflow":[90],"control.":[91],"Motivated":[92],"these":[94],"findings,":[95],"we":[96],"introduce":[97],"VERSE,":[98],"Verified":[100],"Self-Evolving":[101],"harnesses.":[105],"VERSE":[106,169],"lets":[107],"test":[110,167],"draft":[111],"replay":[113],"perturb":[116],"suspected":[117],"steps":[118],"before":[119],"submission,":[120],"tracking":[122],"fixes":[123],"regressions":[125],"across":[126],"rounds.":[127],"Using":[128],"this":[129],"feedback,":[130],"revises":[133],"both":[134],"executor":[136,154],"harness":[137,174,190],"its":[139],"skills,":[142],"hooks,":[144],"notes,":[146],"weights":[149],"models":[155],"stay":[156],"Under":[158],"shared":[160],"protocol":[161],"with":[162],"disjoint":[163],"training,":[164],"validation,":[165],"tasks,":[168],"all":[171],"four":[172],"evaluated":[173],"on":[176],"held-out":[177],"SWE-rebench":[178],"tasks":[179,183],"newer":[181],"out-of-distribution":[182],"in":[184],"languages.":[186],"Its":[187],"validation-selected":[189],"reaches":[191],"42.3%":[192],"37.7%":[194],"accuracy,":[195],"respectively,":[196],"against":[197],"39.2%":[198],"29.3%":[200],"strongest":[203],"baselines.":[204],"Code":[205],"available":[207],"at":[208],"https://github.com/wzekai/VERSE.":[209]},"counts_by_year":[{"year":2026,"cited_by_count":4}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-06T00:00:00"}