{"id":"https://openalex.org/W7214716274","doi":"https://doi.org/10.48550/arxiv.2609.34422","title":"Coding Agent Memory Post-training: Unlocking the Memory Potential of Pre-trained File Operations for Long-Horizon Tasks via Reinforcement Learning","display_name":"Coding Agent Memory Post-training: Unlocking the Memory Potential of Pre-trained File Operations for Long-Horizon Tasks via Reinforcement Learning","publication_year":2026,"publication_date":"2026-09-28","ids":{"openalex":"https://openalex.org/W7214716274","doi":"https://doi.org/10.48550/arxiv.2609.34422"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.34422","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34422","pdf_url":"https://arxiv.org/pdf/2609.34422","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/2609.34422","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5127962369","display_name":"Lirui Luo","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Lirui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074229248","display_name":"Kelong Mao","orcid":"https://orcid.org/0000-0002-5648-568X","observed_orcids":["https://orcid.org/0000-0002-5648-568X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mao, Kelong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103220384","display_name":"Heming Xia","orcid":"https://orcid.org/0000-0001-5074-3441","observed_orcids":["https://orcid.org/0000-0001-5074-3441"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xia, Heming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153453708","display_name":"Rongqing Li","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Rongqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153529053","display_name":"Xinwei Yang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Xinwei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021460597","display_name":"Luyu Chen","orcid":"https://orcid.org/0000-0002-2153-5712","observed_orcids":["https://orcid.org/0000-0002-2153-5712","https://orcid.org/0009-0005-2372-4022"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Luyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153462922","display_name":"Kieran Wong","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wong, Kieran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153453646","display_name":"Yudong Guo","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Yudong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153525037","display_name":"Xinrui Wang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Xinrui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153470215","display_name":"Jiayin Zhu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Jiayin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086564536","display_name":"Simiu Gu","orcid":"https://orcid.org/0000-0002-0113-4540","observed_orcids":["https://orcid.org/0000-0002-0113-4540"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gu, Simiu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153502917","display_name":"Sulong Xu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Sulong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5008843158","display_name":"Cong Fang","orcid":"https://orcid.org/0000-0002-5076-7897","observed_orcids":["https://orcid.org/0000-0002-5076-7897"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fang, Cong","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":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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.49390000104904175,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.49390000104904175,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.1996999979019165,"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/T10260","display_name":"Software Engineering Research","score":0.11320000141859055,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/language-model-agents","display_name":"language model agents","score":0.7149999737739563},{"id":"https://openalex.org/keywords/agent-memory","display_name":"agent memory","score":0.6380000114440918},{"id":"https://openalex.org/keywords/long-horizon-tasks","display_name":"long-horizon tasks","score":0.9559999704360962},{"id":"https://openalex.org/keywords/post-training","display_name":"post-training","score":0.574999988079071},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"reinforcement learning","score":0.5659999847412109},{"id":"https://openalex.org/keywords/qwen3-5","display_name":"Qwen3.5","score":0.8090000152587891},{"id":"https://openalex.org/keywords/swe-bench","display_name":"SWE-bench","score":0.8019999861717224},{"id":"https://openalex.org/keywords/code-generation","display_name":"code generation","score":0.30799999833106995}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8064000010490417},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6887000203132629},{"id":"https://openalex.org/C74426580","wikidata":"https://www.wikidata.org/wiki/Q719484","display_name":"Memory map","level":3,"score":0.532800018787384},{"id":"https://openalex.org/C179518139","wikidata":"https://www.wikidata.org/wiki/Q5140297","display_name":"Coding (social sciences)","level":2,"score":0.4747999906539917},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.46720001101493835},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.46299999952316284},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.4350999891757965},{"id":"https://openalex.org/C160145156","wikidata":"https://www.wikidata.org/wiki/Q778586","display_name":"Executable","level":2,"score":0.43459999561309814},{"id":"https://openalex.org/C136085584","wikidata":"https://www.wikidata.org/wiki/Q910289","display_name":"Overlay","level":2,"score":0.41260001063346863},{"id":"https://openalex.org/C39528615","wikidata":"https://www.wikidata.org/wiki/Q1229610","display_name":"Distributed shared memory","level":5,"score":0.3953000009059906},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.38420000672340393},{"id":"https://openalex.org/C12186640","wikidata":"https://www.wikidata.org/wiki/Q6815743","display_name":"Memory model","level":3,"score":0.3756999969482422},{"id":"https://openalex.org/C57863822","wikidata":"https://www.wikidata.org/wiki/Q905488","display_name":"Flat memory model","level":4,"score":0.37070000171661377},{"id":"https://openalex.org/C76399640","wikidata":"https://www.wikidata.org/wiki/Q189401","display_name":"Virtual memory","level":4,"score":0.36820000410079956},{"id":"https://openalex.org/C171675096","wikidata":"https://www.wikidata.org/wiki/Q1143380","display_name":"Extended memory","level":4,"score":0.35910001397132874},{"id":"https://openalex.org/C197792726","wikidata":"https://www.wikidata.org/wiki/Q191062","display_name":"Mnemonic","level":2,"score":0.3513000011444092},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3400000035762787},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.314300000667572},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.28369998931884766},{"id":"https://openalex.org/C156731835","wikidata":"https://www.wikidata.org/wiki/Q751740","display_name":"Memory leak","level":4,"score":0.2757999897003174},{"id":"https://openalex.org/C51290061","wikidata":"https://www.wikidata.org/wiki/Q1936765","display_name":"Uniform memory access","level":4,"score":0.26989999413490295},{"id":"https://openalex.org/C133875982","wikidata":"https://www.wikidata.org/wiki/Q764810","display_name":"Shared memory","level":2,"score":0.2676999866962433},{"id":"https://openalex.org/C63511323","wikidata":"https://www.wikidata.org/wiki/Q908936","display_name":"Interleaved memory","level":4,"score":0.26600000262260437},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.26570001244544983},{"id":"https://openalex.org/C153247305","wikidata":"https://www.wikidata.org/wiki/Q835713","display_name":"Memory address","level":3,"score":0.26510000228881836},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.2646999955177307},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.2621999979019165},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.2612999975681305},{"id":"https://openalex.org/C176649486","wikidata":"https://www.wikidata.org/wiki/Q2308807","display_name":"Memory management","level":3,"score":0.2603999972343445}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.34422","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34422","pdf_url":"https://arxiv.org/pdf/2609.34422","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.2609.34422","doi":"https://doi.org/10.48550/arxiv.2609.34422","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.34422","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:2609.34422","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34422","pdf_url":"https://arxiv.org/pdf/2609.34422","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/W7214716274.pdf","grobid_xml":"https://content.openalex.org/works/W7214716274.grobid-xml"},"referenced_works_count":23,"referenced_works":["https://openalex.org/W2736601468","https://openalex.org/W4387636003","https://openalex.org/W4414971194","https://openalex.org/W4415276218","https://openalex.org/W4417167303","https://openalex.org/W7116953735","https://openalex.org/W7117734275","https://openalex.org/W7124152780","https://openalex.org/W7154220992","https://openalex.org/W7154426196","https://openalex.org/W7162070983","https://openalex.org/W7162674349","https://openalex.org/W7163005121","https://openalex.org/W7166865363","https://openalex.org/W7167596745","https://openalex.org/W7167815111","https://openalex.org/W7170173822","https://openalex.org/W7171536440","https://openalex.org/W7171849062","https://openalex.org/W7172031919","https://openalex.org/W7172492227","https://openalex.org/W7202167007","https://openalex.org/W7203746987"],"related_works":["https://openalex.org/W7170173822","https://openalex.org/W7166825402","https://openalex.org/W7131825876","https://openalex.org/W7161958096","https://openalex.org/W4417095241","https://openalex.org/W4415270781","https://openalex.org/W7162070983","https://openalex.org/W7124151273","https://openalex.org/W7084138381","https://openalex.org/W7221039799"],"abstract_inverted_index":{"Language-model":[0],"agents":[1,70,117],"increasingly":[2],"tackle":[3],"long-horizon":[4,76,92],"tasks":[5],"whose":[6],"interaction":[7],"histories":[8],"exceed":[9],"the":[10,28,56],"model's":[11,58],"active":[12],"context.":[13],"Recent":[14],"work":[15],"has":[16],"begun":[17],"to":[18,22,50,72,118],"use":[19,73],"reinforcement":[20],"learning":[21,148],"make":[23],"memory":[24,34,48,74,126,151],"control":[25],"part":[26],"of":[27,40,91,167],"policy,":[29],"often":[30],"relying":[31],"on":[32],"predefined":[33],"tools":[35],"within":[36],"domain-specific":[37],"training":[38],"environments":[39,94,143],"relatively":[41],"short":[42],"horizons.":[43],"This":[44],"setup":[45],"ties":[46],"learned":[47,63],"behavior":[49,152],"environment-specific":[51],"interfaces":[52],"that":[53],"lie":[54],"outside":[55],"base":[57],"pre-training":[59],"and":[60,99,112,122,158,162,173,177,183],"must":[61],"be":[62],"from":[64,154,164],"scratch,":[65],"so":[66],"even":[67],"after":[68],"post-training,":[69],"struggle":[71],"in":[75],"tasks.":[77],"To":[78],"address":[79],"these":[80],"limitations,":[81],"we":[82,159],"introduce":[83,132],"Coding":[84],"Agent":[85],"Memory":[86],"Gym":[87],"(CAMG),":[88],"a":[89,136],"suite":[90],"agentic-RL":[93],"spanning":[95],"Shop,":[96],"Coding,":[97],"DeepResearch,":[98],"AutoResearch.":[100],"Alongside":[101],"each":[102],"environment's":[103],"native":[104],"task":[105,156],"interface,":[106],"CAMG":[107],"provides":[108],"executable":[109],"shell":[110],"access":[111],"an":[113,128],"episode-persistent":[114],"workspace,":[115],"enabling":[116],"create,":[119],"revise,":[120],"search,":[121],"reuse":[123],"files":[124],"as":[125],"throughout":[127],"episode.":[129],"We":[130],"also":[131],"CAMG-RL,":[133],"which":[134],"trains":[135],"single":[137],"policy":[138],"jointly":[139],"across":[140],"all":[141],"four":[142],"with":[144,181],"fully":[145],"asynchronous":[146],"PPO,":[147],"this":[149],"file-based":[150],"directly":[153],"downstream":[155],"reward,":[157],"train":[160],"CAMG-RL-4B":[161,176],"CAMG-RL-9B":[163,178],"Qwen3.5":[165],"models":[166],"matching":[168],"size.":[169],"On":[170],"SWE-bench":[171],"Verified":[172],"MLE-bench":[174],"Lite,":[175],"are":[179],"competitive":[180],"Qwen3.5-35B-A3B":[182],"Qwen3.5-122B-A10B,":[184],"respectively.":[185]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-30T00:00:00"}