{"id":"https://openalex.org/W7221772012","doi":"https://doi.org/10.48550/arxiv.2610.18427","title":"Reliable Memory from Unstable Growth: Curating Self-Written Agent Playbooks with Randomized Experiments","display_name":"Reliable Memory from Unstable Growth: Curating Self-Written Agent Playbooks with Randomized Experiments","publication_year":2026,"publication_date":"2026-10-13","ids":{"openalex":"https://openalex.org/W7221772012","doi":"https://doi.org/10.48550/arxiv.2610.18427"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.18427","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.18427","pdf_url":"https://arxiv.org/pdf/2610.18427","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/2610.18427","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5124053802","display_name":"Hoang Pham","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Pham, Hoang","raw_affiliation_strings":["Nanyang Technological University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanyang Technological University","institution_ids":["https://openalex.org/I172675005"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5025645183","display_name":"Cong-Duy T Nguyen","orcid":"https://orcid.org/0000-0002-0931-460X","observed_orcids":["https://orcid.org/0000-0002-0931-460X"]},"institutions":[{"id":"https://openalex.org/I4210142044","display_name":"VinUniversity","ror":"https://ror.org/052dmdr17","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210142044"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Nguyen, Cong-Duy","raw_affiliation_strings":["VinUniversity"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"VinUniversity","institution_ids":["https://openalex.org/I4210142044"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113211568","display_name":"Dong Le Van","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Le, Dong","raw_affiliation_strings":["Nanyang Technological University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanyang Technological University","institution_ids":["https://openalex.org/I172675005"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5106011068","display_name":"Ponhvoan Srey","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Srey, Ponhvoan","raw_affiliation_strings":["Nanyang Technological University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanyang Technological University","institution_ids":["https://openalex.org/I172675005"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5050386762","display_name":"Anh Tuan Luu","orcid":"https://orcid.org/0000-0001-6062-207X","observed_orcids":["https://orcid.org/0000-0001-6062-207X"]},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]},{"id":"https://openalex.org/I4210142044","display_name":"VinUniversity","ror":"https://ror.org/052dmdr17","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210142044"]}],"countries":["SG","VN"],"is_corresponding":false,"raw_author_name":"Luu, Anh Tuan","raw_affiliation_strings":["Nanyang Technological University","VinUniversity"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanyang Technological University","institution_ids":["https://openalex.org/I172675005"]},{"raw_affiliation_string":"VinUniversity","institution_ids":["https://openalex.org/I4210142044"]}]}],"institutions":[],"countries_distinct_count":2,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.8598999977111816,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.8598999977111816,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.017899999395012856,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.01720000058412552,"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/language-model-agents","display_name":"language model agents","score":0.7580000162124634}],"concepts":[{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.7669000029563904},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.7508999705314636},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7304999828338623},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.692300021648407},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5497999787330627},{"id":"https://openalex.org/C155108698","wikidata":"https://www.wikidata.org/wiki/Q1231081","display_name":"Randomized experiment","level":2,"score":0.3953999876976013},{"id":"https://openalex.org/C4679612","wikidata":"https://www.wikidata.org/wiki/Q866298","display_name":"Aggregate (composite)","level":2,"score":0.3675000071525574},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.32580000162124634},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.26649999618530273},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2628999948501587}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.18427","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.18427","pdf_url":"https://arxiv.org/pdf/2610.18427","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.2610.18427","doi":"https://doi.org/10.48550/arxiv.2610.18427","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.18427","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2610.18427","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.18427","pdf_url":"https://arxiv.org/pdf/2610.18427","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":[{"score":0.4156000018119812,"id":"https://metadata.un.org/sdg/4","display_name":"Quality education"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7221772012.pdf","grobid_xml":"https://content.openalex.org/works/W7221772012.grobid-xml"},"referenced_works_count":12,"referenced_works":["https://openalex.org/W3005700362","https://openalex.org/W4297812065","https://openalex.org/W4319049323","https://openalex.org/W4321649030","https://openalex.org/W4387636003","https://openalex.org/W7166692021","https://openalex.org/W7167604932","https://openalex.org/W7171550432","https://openalex.org/W7197014348","https://openalex.org/W7203504004","https://openalex.org/W7204167736","https://openalex.org/W7204175137"],"related_works":["https://openalex.org/W7214199566","https://openalex.org/W7223155727","https://openalex.org/W7221062216","https://openalex.org/W7133508550","https://openalex.org/W7223246107","https://openalex.org/W7135059314","https://openalex.org/W7223233068","https://openalex.org/W7134053239","https://openalex.org/W7166853358","https://openalex.org/W7166802862"],"abstract_inverted_index":{"Persistent":[0],"textual":[1],"memory":[2,204],"lets":[3],"language-model":[4],"agents":[5],"reuse":[6],"lessons":[7,31,50,91,125,236],"from":[8],"past":[9],"tasks":[10,16,150],"to":[11,47,59,81,130,215,237],"improve":[12,51],"on":[13,96,119,135,198],"future":[14],"multi-step":[15],"without":[17,153],"retraining.":[18],"Systems":[19],"such":[20],"as":[21,28],"ACE":[22],"and":[23,38,152,188,208,227],"Dynamic":[24],"Cheatsheet":[25],"keep":[26],"it":[27],"a":[29,72,116,154,180,183],"playbook:":[30],"the":[32,57,60,67,102,158,164,169,175,220,230,235],"agent":[33,61],"writes":[34],"after":[35],"each":[36,136],"task":[37,97],"reads":[39],"in":[40,160],"its":[41],"prompt.":[42],"A":[43],"key":[44],"challenge":[45],"is":[46],"decide":[48],"which":[49,63,111],"later":[52],"tasks.":[53,121],"Current":[54],"systems":[55],"leave":[56],"decision":[58],"itself,":[62],"makes":[64,149],"performance":[65],"unstable:":[66],"same":[68,176],"procedure":[69,177],"can":[70,178],"write":[71,179],"playbook":[73,195,226],"that":[74,78,171],"helps":[75],"or":[76,140,182,193],"one":[77],"falls":[79],"up":[80],"69.6":[82],"points":[83],"below":[84],"no":[85,194,225],"playbook.":[86],"These":[87],"findings":[88],"motivate":[89],"valuing":[90],"by":[92,101],"their":[93,161],"measured":[94],"effect":[95,114],"success":[98,132,162,214],"rather":[99],"than":[100],"agent's":[103,231],"judgment.":[104],"We":[105],"propose":[106],"Randomized":[107],"Block":[108],"Curation":[109],"(RBC),":[110],"measures":[112],"this":[113],"with":[115,151,219,224],"randomized":[117],"experiment":[118],"practice":[120,137,200],"It":[122],"groups":[123],"related":[124],"into":[126],"blocks":[127,170],"large":[128],"enough":[129],"move":[131],"measurably":[133],"and,":[134,173],"task,":[138],"includes":[139],"leaves":[141],"out":[142],"every":[143],"block":[144,155],"at":[145],"random.":[146],"Random":[147],"inclusion":[148],"comparable,":[156],"so":[157],"gap":[159],"estimates":[163],"block's":[165],"effect.":[166],"RBC":[167,211],"keeps":[168],"help":[172],"since":[174],"helpful":[181],"harmful":[184],"playbook,":[185,222],"grows":[186],"several":[187],"serves":[189],"whichever":[190],"curated,":[191],"whole":[192,221],"scores":[196],"best":[197],"fresh":[199],"episodes.":[201],"Across":[202],"two":[203],"systems,":[205],"three":[206,209],"models":[207],"benchmarks,":[210],"raises":[212],"average":[213],"66.9,":[216],"against":[217],"55.5":[218],"54.1":[223],"53.8":[228],"when":[229],"own":[232],"judgment":[233],"picks":[234],"keep.":[238]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-15T00:00:00"}