{"id":"https://openalex.org/W7165643308","doi":"https://doi.org/10.48550/arxiv.2606.22305","title":"Learning at the Right Pace: Adaptive Data Scheduling Improves LLM Reinforcement Learning","display_name":"Learning at the Right Pace: Adaptive Data Scheduling Improves LLM Reinforcement Learning","publication_year":2026,"publication_date":"2026-06-21","ids":{"openalex":"https://openalex.org/W7165643308","doi":"https://doi.org/10.48550/arxiv.2606.22305"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.22305","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.22305","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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.2606.22305","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5073136550","display_name":"Zhewei Xu","orcid":"https://orcid.org/0000-0003-4718-6175"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Zicheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139157056","display_name":"Ruixuan Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Ruixuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139149066","display_name":"Yu-Neng Chuang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chuang, Yu-Neng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139174645","display_name":"Xiuyi Lou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lou, Xiuyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139163799","display_name":"Hoang Anh Duy Le","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Le, Hoang Anh Duy","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139202620","display_name":"Oren Gal","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gal, Oren","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019891335","display_name":"Alexander S. Szalay","orcid":"https://orcid.org/0000-0002-4108-3282"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Szalay, Alexander S.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073136550","display_name":"Zhewei Xu","orcid":"https://orcid.org/0000-0003-4718-6175"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Zhaozhuo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139131620","display_name":"Guanchu Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Guanchu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139165494","display_name":"Vladimir Braverman","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Braverman, Vladimir","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/T10028","display_name":"Topic Modeling","score":0.28139999508857727,"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/T10028","display_name":"Topic Modeling","score":0.28139999508857727,"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/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.0828000009059906,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.07540000230073929,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.7366999983787537},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.704800009727478},{"id":"https://openalex.org/keywords/adaptive-sampling","display_name":"Adaptive sampling","score":0.4106999933719635},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.4050000011920929},{"id":"https://openalex.org/keywords/real-world-data","display_name":"Real world data","score":0.3677999973297119},{"id":"https://openalex.org/keywords/sample","display_name":"Sample (material)","score":0.36070001125335693}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7875000238418579},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7366999983787537},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.704800009727478},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5120000243186951},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4909000098705292},{"id":"https://openalex.org/C2781395549","wikidata":"https://www.wikidata.org/wiki/Q4680762","display_name":"Adaptive sampling","level":3,"score":0.4106999933719635},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.4050000011920929},{"id":"https://openalex.org/C3020493868","wikidata":"https://www.wikidata.org/wiki/Q55631277","display_name":"Real world data","level":2,"score":0.3677999973297119},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.36070001125335693},{"id":"https://openalex.org/C167393769","wikidata":"https://www.wikidata.org/wiki/Q4680768","display_name":"Adaptive strategies","level":2,"score":0.31529998779296875},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.31200000643730164},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.287200003862381},{"id":"https://openalex.org/C58973888","wikidata":"https://www.wikidata.org/wiki/Q1041418","display_name":"Semi-supervised learning","level":2,"score":0.28529998660087585},{"id":"https://openalex.org/C52970973","wikidata":"https://www.wikidata.org/wiki/Q2497134","display_name":"Adaptive system","level":2,"score":0.28060001134872437},{"id":"https://openalex.org/C125014702","wikidata":"https://www.wikidata.org/wiki/Q4680749","display_name":"Adaptive learning","level":2,"score":0.2619999945163727}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.22305","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.22305","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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.2606.22305","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.22305","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"achieve":[4],"remarkable":[5],"reasoning":[6,121],"capabilities":[7],"through":[8],"reinforcement":[9],"learning":[10],"(RL)":[11],"post-training.":[12,159],"However,":[13],"existing":[14],"RL":[15,57,141,158],"post-training":[16,58],"commonly":[17],"relies":[18],"on":[19],"uniform":[20,61],"data":[21,31,52,153],"sampling,":[22],"which":[23,109],"ignores":[24],"the":[25,29,33,37,75,97],"semantic":[26,68,83],"structure":[27],"of":[28,36],"training":[30,38,94],"and":[32,70,85,119],"changing":[34],"capability":[35],"policy.":[39],"To":[40],"address":[41],"these":[42],"limitations,":[43],"we":[44],"propose":[45],"Adaptive":[46],"Data":[47],"Scheduling":[48],"(ADS),":[49],"a":[50,151],"dual-level":[51],"scheduling":[53,103,154],"framework":[54],"for":[55,156],"pacing":[56],"that":[59,124],"replaces":[60],"sampling":[62],"with":[63,143],"an":[64,87],"adaptive":[65,88],"distribution":[66,90],"over":[67,131],"clusters":[69],"policy-boundary":[71,107],"sample":[72,98,106],"selection.":[73],"At":[74,96],"cluster":[76],"level,":[77,99],"ADS":[78,100,125,138],"organizes":[79],"samples":[80],"according":[81],"to":[82,91,104],"patterns":[84],"maintains":[86],"inter-cluster":[89],"solidify":[92],"current":[93],"progress.":[95],"performs":[101],"intra-cluster":[102],"continuously":[105],"samples,":[108],"provides":[110],"informative":[111],"relative":[112],"advantages.":[113],"Experimental":[114],"results":[115],"across":[116],"three":[117],"LLMs":[118],"seven":[120],"benchmarks":[122],"demonstrate":[123],"improves":[126,140],"average":[127],"accuracy":[128],"by":[129],"5.2%":[130],"Group":[132],"Relative":[133],"Policy":[134],"Optimization":[135],"(GRPO).":[136],"Notably,":[137],"consistently":[139],"methods":[142],"different":[144],"objective":[145],"designs,":[146],"highlighting":[147],"its":[148],"potential":[149],"as":[150],"general":[152],"strategy":[155],"LLM":[157],"The":[160],"source":[161],"code":[162],"is":[163],"available":[164],"at:":[165],"https://github.com/Richard-zrx/ADS.":[166]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-24T00:00:00"}