{"id":"https://openalex.org/W7202265054","doi":"https://doi.org/10.48550/arxiv.2608.10424","title":"Recovering Wasted Compute in Autoresearch Agents","display_name":"Recovering Wasted Compute in Autoresearch Agents","publication_year":2026,"publication_date":"2026-08-11","ids":{"openalex":"https://openalex.org/W7202265054","doi":"https://doi.org/10.48550/arxiv.2608.10424"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.10424","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.10424","pdf_url":"https://arxiv.org/pdf/2608.10424","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":"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/2608.10424","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5147060925","display_name":"Au Kwok Chun","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chun, Au Kwok","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147011544","display_name":"Abhigyan Acherjee","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I184565670","display_name":"Georgetown University","ror":"https://ror.org/05vzafd60","country_code":"US","type":"education","lineage":["https://openalex.org/I184565670"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Acherjee, Abhigyan","raw_affiliation_strings":["AI, Analytics and Future of Work Initiative , Georgetown University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"AI, Analytics and Future of Work Initiative , Georgetown University","institution_ids":["https://openalex.org/I184565670"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5146950473","display_name":"Amrutha Rao","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I78577930","display_name":"Columbia University","ror":"https://ror.org/00hj8s172","country_code":"US","type":"education","lineage":["https://openalex.org/I78577930"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Rao, Amrutha","raw_affiliation_strings":["Department of Applied Mathematics , Columbia University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Applied Mathematics , Columbia University","institution_ids":["https://openalex.org/I78577930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5146949493","display_name":"Zaiqian Chen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I78577930","display_name":"Columbia University","ror":"https://ror.org/00hj8s172","country_code":"US","type":"education","lineage":["https://openalex.org/I78577930"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chen, Zaiqian","raw_affiliation_strings":["Department of Statistics , Columbia University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Statistics , Columbia University","institution_ids":["https://openalex.org/I78577930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029940599","display_name":"Kazem Meidani","orcid":"https://orcid.org/0000-0001-8441-8700","observed_orcids":["https://orcid.org/0000-0001-8441-8700"]},"institutions":[{"id":"https://openalex.org/I1305444813","display_name":"Capital One (United States)","ror":"https://ror.org/00svp7168","country_code":"US","type":"company","lineage":["https://openalex.org/I1305444813"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Meidani, Kazem","raw_affiliation_strings":["Capital One"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Capital One","institution_ids":["https://openalex.org/I1305444813"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017427561","display_name":"C. Bayan Bruss","orcid":"https://orcid.org/0009-0006-2260-7518","observed_orcids":["https://orcid.org/0009-0006-2260-7518"]},"institutions":[{"id":"https://openalex.org/I1305444813","display_name":"Capital One (United States)","ror":"https://ror.org/00svp7168","country_code":"US","type":"company","lineage":["https://openalex.org/I1305444813"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bruss, C. Bayan","raw_affiliation_strings":["Capital One"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Capital One","institution_ids":["https://openalex.org/I1305444813"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5066564672","display_name":"Micah Goldblum","orcid":"https://orcid.org/0000-0002-8266-2424","observed_orcids":["https://orcid.org/0000-0002-8266-2424"]},"institutions":[{"id":"https://openalex.org/I78577930","display_name":"Columbia University","ror":"https://ror.org/00hj8s172","country_code":"US","type":"education","lineage":["https://openalex.org/I78577930"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Goldblum, Micah","raw_affiliation_strings":["Department of Computer Science , Columbia University","Department of Electrical Engineering , Columbia University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Columbia University","institution_ids":["https://openalex.org/I78577930"]},{"raw_affiliation_string":"Department of Electrical Engineering , Columbia University","institution_ids":["https://openalex.org/I78577930"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.7875999808311462,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.7875999808311462,"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/T11975","display_name":"Evolutionary Algorithms and Applications","score":0.034699998795986176,"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/T11986","display_name":"Scientific Computing and Data Management","score":0.0340999998152256,"subfield":{"id":"https://openalex.org/subfields/1802","display_name":"Information Systems and Management"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/tabular-data-analysis","display_name":"tabular data analysis","score":0.5950000286102295},{"id":"https://openalex.org/keywords/hyperparameter-optimization","display_name":"hyperparameter optimization","score":0.8149999976158142},{"id":"https://openalex.org/keywords/tree-search","display_name":"tree search","score":0.531000018119812},{"id":"https://openalex.org/keywords/prompt-engineering","display_name":"prompt engineering","score":0.8610000014305115},{"id":"https://openalex.org/keywords/failure-modes","display_name":"failure modes","score":0.5239999890327454},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.6399999856948853}],"concepts":[{"id":"https://openalex.org/C168065819","wikidata":"https://www.wikidata.org/wiki/Q845566","display_name":"Debugging","level":2,"score":0.7554000020027161},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6909999847412109},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.6331999897956848},{"id":"https://openalex.org/C2776207758","wikidata":"https://www.wikidata.org/wiki/Q5303302","display_name":"Downstream (manufacturing)","level":2,"score":0.43810001015663147},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.4212000072002411},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4081000089645386},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40149998664855957},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.36309999227523804},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.35269999504089355},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.3398999869823456},{"id":"https://openalex.org/C3020493868","wikidata":"https://www.wikidata.org/wiki/Q55631277","display_name":"Real world data","level":2,"score":0.3181000053882599},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2962000072002411},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.28940001130104065},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.287200003862381},{"id":"https://openalex.org/C124304363","wikidata":"https://www.wikidata.org/wiki/Q673661","display_name":"Abstraction","level":2,"score":0.2574000060558319},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.25279998779296875}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.10424","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.10424","pdf_url":"https://arxiv.org/pdf/2608.10424","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.10424","doi":"https://doi.org/10.48550/arxiv.2608.10424","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.10424","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:2608.10424","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.10424","pdf_url":"https://arxiv.org/pdf/2608.10424","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7202265054.pdf","grobid_xml":"https://content.openalex.org/works/W7202265054.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7169827629","https://openalex.org/W7220897681","https://openalex.org/W7203726719","https://openalex.org/W7203597325","https://openalex.org/W7214656245","https://openalex.org/W7164365481","https://openalex.org/W7212233845","https://openalex.org/W7161254174","https://openalex.org/W7202124398","https://openalex.org/W7160635469"],"abstract_inverted_index":{"A":[0],"slew":[1],"of":[2,54,152],"recent":[3],"works":[4],"develop":[5],"agents":[6,20],"for":[7,40],"solving":[8],"research":[9],"problems":[10],"end-to-end,":[11],"a":[12,93,140],"paradigm":[13],"increasingly":[14],"referred":[15],"to":[16,30,67,86,129],"as":[17],"autoresearch.":[18],"Such":[19],"have":[21,92],"inspired":[22],"large":[23,94,172],"industry":[24],"investment,":[25],"motivated":[26],"by":[27],"their":[28],"potential":[29],"automate":[31],"time-consuming":[32],"human":[33],"labor":[34],"and":[35,58,79,108,137,157,160],"customize":[36],"machine":[37],"learning":[38],"solutions":[39],"specialized":[41],"applications.":[42],"In":[43],"this":[44],"paper,":[45],"we":[46],"study":[47],"the":[48,52,75,99,115,153,185],"modeling":[49],"pipeline":[50],"at":[51],"core":[53],"these":[55],"autoresearch":[56,175],"systems":[57],"identify":[59],"common":[60],"failure":[61],"modes":[62],"when":[63,90],"they":[64,71,83,91,110,119],"are":[65,120,178],"applied":[66],"tabular":[68],"datasets:":[69],"(1)":[70],"waste":[72],"compute":[73,96],"resolving":[74],"same":[76],"bugs":[77],"over":[78,80],"again;":[81],"(2)":[82],"often":[84],"fail":[85],"tune":[87],"hyperparameters":[88],"even":[89],"remaining":[95],"budget;":[97],"(3)":[98],"tree-search":[100,162],"algorithms":[101,163],"that":[102,127,139,144,171],"power":[103],"them":[104],"do":[105,124],"not":[106,125],"explore;":[107],"(4)":[109],"perform":[111],"data":[112,118],"analysis,":[113],"mimicking":[114],"humans":[116],"whose":[117],"trained":[121],"on,":[122],"but":[123],"use":[126],"analysis":[128],"make":[130],"downstream":[131],"decisions.":[132],"We":[133],"explore":[134],"targeted":[135],"interventions":[136],"find":[138],"global":[141],"debug":[142],"consultant":[143],"shares":[145],"discovered":[146],"runtime":[147],"constraints":[148],"across":[149],"all":[150],"branches":[151],"search":[154],"tree,":[155],"prompt-":[156],"control-level":[158],"enhancements,":[159],"refined":[161],"successfully":[164],"recover":[165],"wasted":[166],"compute.":[167],"Our":[168],"results":[169],"show":[170],"gains":[173],"in":[174],"agent":[176],"performance":[177],"achievable":[179],"through":[180],"agentic":[181],"design":[182],"alone,":[183],"holding":[184],"underlying":[186],"language":[187],"model":[188],"fixed.":[189]},"counts_by_year":[],"updated_date":"2026-10-29T09:56:32.521200","created_date":"2026-08-13T00:00:00"}