{"id":"https://openalex.org/W7160824624","doi":"https://doi.org/10.48550/arxiv.2605.07063","title":"Dr. Post-Training: A Data Regularization Perspective on LLM Post-Training","display_name":"Dr. Post-Training: A Data Regularization Perspective on LLM Post-Training","publication_year":2026,"publication_date":"2026-05-08","ids":{"openalex":"https://openalex.org/W7160824624","doi":"https://doi.org/10.48550/arxiv.2605.07063"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.07063","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07063","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":"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.07063","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5030390478","display_name":"Pingbang Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Pingbang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135904654","display_name":"Xueshen Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Xueshen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135849319","display_name":"Z. Morley Mao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mao, Z. Morley","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135857807","display_name":"Jiaqi W. Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Jiaqi W.","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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.3382999897003174,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.3382999897003174,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.1589999943971634,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.11299999803304672,"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/overfitting","display_name":"Overfitting","score":0.7907000184059143},{"id":"https://openalex.org/keywords/regularization","display_name":"Regularization (linguistics)","score":0.5304999947547913},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.5152000188827515},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.41920000314712524},{"id":"https://openalex.org/keywords/framing","display_name":"Framing (construction)","score":0.40130001306533813},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.3887999951839447},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.37940001487731934}],"concepts":[{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.7907000184059143},{"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/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5636000037193298},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.5304999947547913},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.5152000188827515},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.44769999384880066},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.41920000314712524},{"id":"https://openalex.org/C169087156","wikidata":"https://www.wikidata.org/wiki/Q2131593","display_name":"Framing (construction)","level":2,"score":0.40130001306533813},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.39750000834465027},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.3887999951839447},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.37940001487731934},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.35269999504089355},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.35089999437332153},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.31690001487731934},{"id":"https://openalex.org/C58489278","wikidata":"https://www.wikidata.org/wiki/Q1172284","display_name":"Data set","level":2,"score":0.3100000023841858},{"id":"https://openalex.org/C2988382989","wikidata":"https://www.wikidata.org/wiki/Q370685","display_name":"Data space","level":2,"score":0.3061000108718872},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.2992999851703644},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2799000144004822},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.26489999890327454},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.2563000023365021},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.2547000050544739}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.07063","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07063","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.07063","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07063","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":"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Data":[0],"selection":[1,110,189],"methods":[2,111,125,147,170,184],"address":[3],"a":[4,39,48,63,77,131,144,149],"critical":[5],"challenge":[6],"in":[7],"LLM":[8],"post-training:":[9],"effectively":[10],"leveraging":[11],"scarce,":[12],"high-fidelity":[13],"target":[14,57,99],"data":[15,46,100,109,188],"alongside":[16],"abundant":[17],"but":[18],"imperfectly":[19],"aligned":[20],"general":[21,44,86],"training":[22,45,74,87,106],"data.":[23],"In":[24],"this":[25,140],"work,":[26],"we":[27,142,162],"move":[28],"beyond":[29],"the":[30,55,85,91,97,120],"data-selection":[31],"framing":[32],"and":[33,89,107,123,153,179,191],"introduce":[34,163],"Dr.":[35],"Post-Training":[36],"(Data-Regularized":[37],"Post-Training),":[38],"novel":[40],"framework":[41,69],"that":[42,51,71,102,167,182],"reconceptualizes":[43],"as":[47,62,113],"data-induced":[49,121],"regularizer":[50],"prevents":[52],"overfitting":[53],"to":[54,127],"scarce":[56,98],"objective,":[58],"rather":[59],"than":[60],"serving":[61],"pool":[64],"for":[65],"selection.":[66],"Specifically,":[67],"our":[68,183],"proposes":[70],"at":[72],"each":[73],"step,":[75],"construct":[76],"feasible":[78,103],"set":[79],"of":[80,119,146],"model":[81,92],"update":[82,93],"directions":[83],"using":[84],"data,":[88],"project":[90],"direction":[94],"specified":[95],"by":[96],"onto":[101],"set.":[104],"Standard":[105],"existing":[108],"arise":[112],"special":[114],"cases":[115],"with":[116,134,171],"different":[117,128,135],"choices":[118],"regularizer,":[122],"these":[124,169],"correspond":[126],"points":[129],"on":[130,139],"bias--variance":[132,156],"spectrum":[133],"regularization":[136],"strength.":[137],"Building":[138],"view,":[141],"propose":[143],"family":[145],"offering":[148],"richer":[150],"design":[151],"space":[152],"more":[154],"flexible":[155],"tradeoffs.":[157],"For":[158],"practical":[159],"LLM-scale":[160],"use,":[161],"careful":[164],"system":[165,192],"optimizations":[166],"realize":[168],"minimal":[172],"overhead.":[173],"Extensive":[174],"experiments":[175],"across":[176],"SFT,":[177],"RLHF,":[178],"RLVR":[180],"show":[181],"consistently":[185],"outperform":[186],"state-of-the-art":[187],"baselines,":[190],"benchmarks":[193],"confirm":[194],"their":[195],"efficiency.":[196]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-12T00:00:00"}