{"id":"https://openalex.org/W7162554468","doi":"https://doi.org/10.48550/arxiv.2605.26275","title":"SPEAR: Code-Augmented Agentic Prompt Optimization","display_name":"SPEAR: Code-Augmented Agentic Prompt Optimization","publication_year":2026,"publication_date":"2026-05-25","ids":{"openalex":"https://openalex.org/W7162554468","doi":"https://doi.org/10.48550/arxiv.2605.26275"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.26275","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.26275","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,"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.2605.26275","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5001674861","display_name":"Mengyin Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Mengyin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137103555","display_name":"Cong Feng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Feng, Cong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137177980","display_name":"Huimin Han","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Huimin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137176545","display_name":"Guangming Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Guangming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137086986","display_name":"Yu Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Yu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137188193","display_name":"Xiaonan Ding","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ding, Xiaonan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137095900","display_name":"Shihui Long","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Long, Shihui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111175142","display_name":"Fengyi Li","orcid":"https://orcid.org/0000-0001-9943-0091"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Fengyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137179541","display_name":"Tanvi Motwani","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Motwani, Tanvi","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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.15479999780654907,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.15479999780654907,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/T12535","display_name":"Machine Learning and Data Classification","score":0.10610000044107437,"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":0.05139999836683273,"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/python","display_name":"Python (programming language)","score":0.7217000126838684},{"id":"https://openalex.org/keywords/confusion","display_name":"Confusion","score":0.6183000206947327},{"id":"https://openalex.org/keywords/lever","display_name":"Lever","score":0.4941999912261963},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.4235999882221222},{"id":"https://openalex.org/keywords/classifier","display_name":"Classifier (UML)","score":0.3993000090122223},{"id":"https://openalex.org/keywords/guard","display_name":"Guard (computer science)","score":0.3619999885559082}],"concepts":[{"id":"https://openalex.org/C519991488","wikidata":"https://www.wikidata.org/wiki/Q28865","display_name":"Python (programming language)","level":2,"score":0.7217000126838684},{"id":"https://openalex.org/C2781140086","wikidata":"https://www.wikidata.org/wiki/Q557945","display_name":"Confusion","level":2,"score":0.6183000206947327},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6055999994277954},{"id":"https://openalex.org/C107524782","wikidata":"https://www.wikidata.org/wiki/Q40164","display_name":"Lever","level":2,"score":0.4941999912261963},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.4235999882221222},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.3993000090122223},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.36340001225471497},{"id":"https://openalex.org/C141141315","wikidata":"https://www.wikidata.org/wiki/Q2379942","display_name":"Guard (computer science)","level":2,"score":0.3619999885559082},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3425999879837036},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.3375999927520752},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.31060001254081726},{"id":"https://openalex.org/C4924752","wikidata":"https://www.wikidata.org/wiki/Q184148","display_name":"Plug-in","level":2,"score":0.2865999937057495},{"id":"https://openalex.org/C52146309","wikidata":"https://www.wikidata.org/wiki/Q7431116","display_name":"Schema (genetic algorithms)","level":2,"score":0.2786000072956085},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.27070000767707825},{"id":"https://openalex.org/C198647972","wikidata":"https://www.wikidata.org/wiki/Q44475","display_name":"Spear","level":2,"score":0.27070000767707825}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.26275","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.26275","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,"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.2605.26275","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.26275","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,"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":{"Automatic":[0],"prompt":[1],"engineering":[2],"(APE)":[3],"rewrites":[4],"prompts":[5],"to":[6,34,64],"improve":[7],"downstream":[8],"task":[9,147],"performance,":[10],"but":[11],"existing":[12],"APE":[13,35],"loops":[14],"treat":[15],"the":[16,25,71,74,82,97,104,149,169,187,191,202,209,230],"optimizer":[17,48,75],"itself":[18,99],"as":[19],"a":[20,45,108,223],"fixed":[21],"pipeline.":[22],"We":[23,121],"port":[24],"code-as-action":[26],"paradigm":[27],"of":[28],"CodeAct":[29],"(Wang":[30],"et":[31],"al.,":[32],"2024a)":[33],"and":[36,62,77,115,134,141,182],"propose":[37],"SPEAR":[38,143,175],"(Sandboxed":[39],"Prompt":[40],"Engineer":[41],"with":[42,49],"Active":[43],"Roll-back),":[44],"free-form":[46],"agentic":[47],"four":[50],"tools":[51],"--":[52,57],"evaluate,":[53],"python,":[54],"set_prompt,":[55],"finish":[56],"that":[58,222],"decides":[59],"autonomously":[60],"how":[61],"when":[63,213],"use":[65],"them.":[66],"The":[67],"distinctive":[68],"tool":[69,189],"is":[70,190,218],"Python":[72,80,188],"sandbox:":[73],"writes":[76],"executes":[78],"arbitrary":[79],"on":[81,112,123,148,156,162,168,195,201,208],"current":[83],"evaluation":[84],"DataFrame,":[85],"performing":[86],"structural":[87],"error":[88,92],"analysis":[89],"(confusion":[90],"matrices,":[91],"clustering,":[93],"per":[94],"group":[95],"metrics)":[96],"agent":[98,106],"authors.":[100],"Two":[101],"guardrails":[102],"turn":[103],"long-horizon":[105],"into":[107],"monotone-improving":[109],"optimizer:":[110],"auto-rollback":[111],"metric":[113,119,151],"regression,":[114],"an":[116],"optional":[117],"guard":[118],"floor.":[120],"evaluate":[122],"three":[124],"industrial":[125,146],"LLM-as-judge":[126],"suites":[127],"(13":[128],"judge":[129,197],"tasks":[130,140,198],"across":[131],"recruiter-intake,":[132],"conversational-memory,":[133],"query-refinement":[135],"systems)":[136],"plus":[137],"seven":[138],"BBH":[139],"GSM8K.":[142],"wins":[144],"every":[145],"primary":[150],"($\u03ba$":[152],"0.857":[153],"vs":[154,160,166,179],"0.359":[155],"tool-selection;":[157],"F1-macro":[158],"0.815":[159],"0.763":[161],"filter-relevance;":[163],"$\u03ba$":[164],"0.254":[165],"0.218":[167],"hardest":[170,210],"extraction":[171,211],"dimension).":[172],"On":[173],"BBH-7":[174],"averages":[176],"0.938":[177],"accuracy":[178],"GEPA":[180],"0.628":[181],"TextGrad":[183],"0.484.":[184],"Ablations":[185],"show":[186],"largest":[192],"single":[193],"lever":[194],"complex":[196],"($\u0394\\approx":[199],"+0.79\u03ba$":[200],"5-class":[203],"tool-selection":[204],"judge,":[205],"$\u0394\\approx":[206],"+0.35\u03ba$":[207],"dimension":[212],"removed);":[214],"its":[215],"irreplaceable":[216],"contribution":[217],"class-pair":[219],"confusion":[220],"aggregation":[221],"long-context":[224],"LLM":[225],"cannot":[226],"extract":[227],"reliably":[228],"from":[229],"raw":[231],"eval":[232],"DataFrame.":[233]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-28T00:00:00"}