{"id":"https://openalex.org/W7224117237","doi":"https://doi.org/10.48550/arxiv.2610.36589","title":"InfraDev: Can Coding Agents Develop AI Infrastructure, and How Can We Tell?","display_name":"InfraDev: Can Coding Agents Develop AI Infrastructure, and How Can We Tell?","publication_year":2026,"publication_date":"2026-10-27","ids":{"openalex":"https://openalex.org/W7224117237","doi":"https://doi.org/10.48550/arxiv.2610.36589"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.36589","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.36589","pdf_url":"https://arxiv.org/pdf/2610.36589","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/2610.36589","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5002898613","display_name":"Yangshen Deng","orcid":"https://orcid.org/0009-0003-9487-1455","observed_orcids":["https://orcid.org/0009-0003-9487-1455"]},"institutions":[{"id":"https://openalex.org/I98677209","display_name":"University of Edinburgh","ror":"https://ror.org/01nrxwf90","country_code":"GB","type":"education","lineage":["https://openalex.org/I98677209"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Deng, Yangshen","raw_affiliation_strings":["University of Edinburgh"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Edinburgh","institution_ids":["https://openalex.org/I98677209"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078324582","display_name":"Jiaqi Ruan","orcid":"https://orcid.org/0000-0003-2584-0738","observed_orcids":["https://orcid.org/0000-0003-2584-0738"]},"institutions":[{"id":"https://openalex.org/I98677209","display_name":"University of Edinburgh","ror":"https://ror.org/01nrxwf90","country_code":"GB","type":"education","lineage":["https://openalex.org/I98677209"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Ruan, Jiaqi","raw_affiliation_strings":["University of Edinburgh"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Edinburgh","institution_ids":["https://openalex.org/I98677209"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160032168","display_name":"Yeqi Huang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98677209","display_name":"University of Edinburgh","ror":"https://ror.org/01nrxwf90","country_code":"GB","type":"education","lineage":["https://openalex.org/I98677209"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Huang, Yeqi","raw_affiliation_strings":["University of Edinburgh"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Edinburgh","institution_ids":["https://openalex.org/I98677209"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5159991760","display_name":"Nick Lau","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I2250653659","display_name":"Tencent (China)","ror":"https://ror.org/00hhjss72","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250653659"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lau, Nick","raw_affiliation_strings":["Tencent Hunyuan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tencent Hunyuan","institution_ids":["https://openalex.org/I2250653659"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020661868","display_name":"Guangming Sheng","orcid":"https://orcid.org/0000-0003-3395-3994","observed_orcids":["https://orcid.org/0000-0003-3395-3994"]},"institutions":[{"id":"https://openalex.org/I2250653659","display_name":"Tencent (China)","ror":"https://ror.org/00hhjss72","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250653659"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Sheng, Guangming","raw_affiliation_strings":["Tencent Hunyuan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tencent Hunyuan","institution_ids":["https://openalex.org/I2250653659"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101924633","display_name":"Luo Mai","orcid":"https://orcid.org/0000-0002-7712-7558","observed_orcids":["https://orcid.org/0000-0002-7712-7558"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mai, Luo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"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/T10260","display_name":"Software Engineering Research","score":0.9491999745368958,"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":0.9491999745368958,"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/T10743","display_name":"Software Testing and Debugging Techniques","score":0.03150000050663948,"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"}},{"id":"https://openalex.org/T13904","display_name":"Artificial Intelligence Applications","score":0.003700000001117587,"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/ai-coding-agents","display_name":"coding agents","score":0.9760000109672546},{"id":"https://openalex.org/keywords/ai-infrastructure","display_name":"AI infrastructure","score":0.7200000286102295},{"id":"https://openalex.org/keywords/benchmarking","display_name":"benchmarking","score":0.5789999961853027},{"id":"https://openalex.org/keywords/pull-requests","display_name":"pull requests","score":0.45100000500679016},{"id":"https://openalex.org/keywords/vllm","display_name":"vLLM","score":0.9990000128746033},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.625},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"reinforcement learning","score":0.460999995470047},{"id":"https://openalex.org/keywords/inference","display_name":"inference","score":0.33500000834465027},{"id":"https://openalex.org/keywords/failure-analysis","display_name":"failure analysis","score":0.6380000114440918}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7965999841690063},{"id":"https://openalex.org/C179518139","wikidata":"https://www.wikidata.org/wiki/Q5140297","display_name":"Coding (social sciences)","level":2,"score":0.6186000108718872},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5396999716758728},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5177000164985657},{"id":"https://openalex.org/C190839683","wikidata":"https://www.wikidata.org/wiki/Q2448197","display_name":"Train","level":2,"score":0.45210000872612},{"id":"https://openalex.org/C2778571376","wikidata":"https://www.wikidata.org/wiki/Q1355821","display_name":"Frontier","level":2,"score":0.42480000853538513},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4212000072002411},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4187000095844269},{"id":"https://openalex.org/C2780428219","wikidata":"https://www.wikidata.org/wiki/Q16952335","display_name":"Cover (algebra)","level":2,"score":0.4016000032424927},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.38190001249313354},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3292999863624573},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.32710000872612},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.29510000348091125},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2770000100135803},{"id":"https://openalex.org/C167822520","wikidata":"https://www.wikidata.org/wiki/Q176452","display_name":"Finite-state machine","level":2,"score":0.2662000060081482}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.36589","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.36589","pdf_url":"https://arxiv.org/pdf/2610.36589","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.2610.36589","doi":"https://doi.org/10.48550/arxiv.2610.36589","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.36589","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:2610.36589","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.36589","pdf_url":"https://arxiv.org/pdf/2610.36589","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":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7224117237.pdf","grobid_xml":"https://content.openalex.org/works/W7224117237.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7127269374","https://openalex.org/W7131417132","https://openalex.org/W7212141264","https://openalex.org/W7161724140","https://openalex.org/W7140307387","https://openalex.org/W7124766172","https://openalex.org/W7202393689","https://openalex.org/W7165399465","https://openalex.org/W7221108603","https://openalex.org/W7132852810"],"abstract_inverted_index":{"Coding":[0],"agents":[1],"are":[2],"beginning":[3],"to":[4,52,71],"develop":[5],"the":[6,98,100,112,117,150,156,158,163,172,213,221],"production-grade":[7],"AI":[8],"infrastructure":[9,58],"that":[10,32,76,108,126,165,188,195],"trains":[11],"and":[12,26,30,36,55,63,80,105,129,137,144,170,190,194],"serves":[13],"large":[14],"models,":[15,149],"yet":[16],"existing":[17],"benchmarks":[18],"cover":[19],"only":[20,153],"single":[21,183],"kernels":[22],"or":[23,203],"single-GPU":[24],"tasks,":[25,132],"grade":[27,88],"with":[28,171],"tests":[29,102,198],"prompts":[31],"misjudge":[33],"both":[34],"correct":[35],"incorrect":[37],"patches.":[38],"We":[39,119],"introduce":[40],"InfraDev,":[41],"a":[42,73,106,122,182,217],"benchmark":[43],"of":[44,155,162,216],"36":[45],"tasks":[46,223],"built":[47],"from":[48],"real-world":[49],"pull":[50,174],"requests":[51],"Megatron-LM,":[53],"vLLM,":[54],"verl,":[56],"covering":[57],"development":[59],"for":[60],"training,":[61],"inference,":[62],"reinforcement":[64],"learning.":[65],"Each":[66],"task":[67],"asks":[68],"an":[69,209],"agent":[70,210],"reimplement":[72],"real":[74],"change":[75],"spans":[77],"many":[78,85],"files":[79],"must":[81],"run":[82],"correctly":[83],"across":[84,131],"GPUs.":[86],"To":[87],"such":[89],"changes,":[90],"we":[91],"jointly":[92],"synthesize":[93],"hidden":[94],"probes,":[95],"which":[96],"enforce":[97],"requirements":[99],"original":[101,173],"leave":[103],"unchecked,":[104],"prompt":[107],"states":[109],"every":[110,167],"interface":[111],"probes":[113,159],"call":[114],"without":[115],"revealing":[116],"solution.":[118],"also":[120],"design":[121],"shared":[123],"execution":[124],"environment":[125],"reuses":[127],"dependencies":[128],"artifacts":[130],"cutting":[133],"storage":[134],"by":[135,140,224],"52-57%":[136],"build":[138],"time":[139],"13-14%":[141],"on":[142,220],"Megatron-LM":[143,222],"vLLM.":[145],"Across":[146],"six":[147],"frontier":[148,218],"best":[151],"resolves":[152,181],"38.9%":[154],"tasks;":[157],"reject":[160],"32%":[161],"submissions":[164],"pass":[166],"official":[168],"test,":[169],"request":[175],"descriptions":[176],"as":[177],"prompts,":[178],"no":[179],"model":[180,219],"task.":[184],"Failure":[185],"analysis":[186],"shows":[187],"structural":[189],"state":[191],"errors":[192],"dominate":[193],"agents'":[196],"own":[197],"rarely":[199],"check":[200],"memory,":[201],"time,":[202],"traffic.":[204],"Turning":[205],"these":[206],"findings":[207],"into":[208],"skill":[211],"raises":[212],"resolved":[214],"rate":[215],"10.5%.":[225]},"counts_by_year":[],"updated_date":"2026-11-06T12:05:46.655746","created_date":"2026-10-29T00:00:00"}