{"id":"https://openalex.org/W4415317404","doi":"https://doi.org/10.48550/arxiv.2510.06826","title":"Mid-Training of Large Language Models: A Survey","display_name":"Mid-Training of Large Language Models: A Survey","publication_year":2025,"publication_date":"2025-10-08","ids":{"openalex":"https://openalex.org/W4415317404","doi":"https://doi.org/10.48550/arxiv.2510.06826"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2510.06826","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.06826","pdf_url":"https://arxiv.org/pdf/2510.06826","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","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/2510.06826","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5048924349","display_name":"Kaixiang Mo","orcid":"https://orcid.org/0000-0003-3326-8938","observed_orcids":["https://orcid.org/0000-0003-3326-8938"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mo, Kaixiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071070368","display_name":"Yuxin Shi","orcid":"https://orcid.org/0000-0001-8529-6549","observed_orcids":["https://orcid.org/0000-0001-8529-6549"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shi, Yuxin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Weng, Weiwei","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Weng, Weiwei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033336832","display_name":"Zhiqiang Zhou","orcid":"https://orcid.org/0000-0001-7436-231X","observed_orcids":["https://orcid.org/0000-0001-7436-231X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Zhiqiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020189532","display_name":"Shuman Liu","orcid":"https://orcid.org/0000-0002-9470-8565","observed_orcids":["https://orcid.org/0000-0002-9470-8565"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Shuman","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zhang, Haibo","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Haibo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5076875538","display_name":"Anxiang Zeng","orcid":"https://orcid.org/0000-0003-3869-5357","observed_orcids":["https://orcid.org/0000-0003-3869-5357"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zeng, Anxiang","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":7,"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":0.9839000105857849,"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":0.9839000105857849,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.00930000003427267,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.0034000000450760126,"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/large-language-models","display_name":"large language models","score":0.9190000295639038},{"id":"https://openalex.org/keywords/fine-tuning","display_name":"fine-tuning","score":0.7649999856948853},{"id":"https://openalex.org/keywords/learning-rate-schedules","display_name":"learning rate scheduling","score":0.7319999933242798},{"id":"https://openalex.org/keywords/gradient-noise","display_name":"gradient noise","score":0.7570000290870667},{"id":"https://openalex.org/keywords/information-bottleneck","display_name":"information bottleneck","score":0.9959999918937683},{"id":"https://openalex.org/keywords/curriculum-learning","display_name":"curriculum learning","score":0.9990000128746033},{"id":"https://openalex.org/keywords/data-quality","display_name":"data quality","score":0.5090000033378601},{"id":"https://openalex.org/keywords/convergence-stability","display_name":"convergence stability","score":0.5180000066757202},{"id":"https://openalex.org/keywords/generalization","display_name":"generalization","score":0.75}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7250999808311462},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.498199999332428},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.49619999527931213},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.49459999799728394},{"id":"https://openalex.org/C58642233","wikidata":"https://www.wikidata.org/wiki/Q8269924","display_name":"Taxonomy (biology)","level":2,"score":0.4909999966621399},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46219998598098755},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.4099999964237213},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4072999954223633},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.35370001196861267},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3310000002384186},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.30399999022483826},{"id":"https://openalex.org/C24756922","wikidata":"https://www.wikidata.org/wiki/Q1757694","display_name":"Data quality","level":3,"score":0.2913999855518341},{"id":"https://openalex.org/C2778464652","wikidata":"https://www.wikidata.org/wiki/Q309849","display_name":"Open research","level":2,"score":0.290800005197525},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.28299999237060547},{"id":"https://openalex.org/C169590947","wikidata":"https://www.wikidata.org/wiki/Q47506","display_name":"Compiler","level":2,"score":0.2680000066757202},{"id":"https://openalex.org/C74672266","wikidata":"https://www.wikidata.org/wiki/Q815859","display_name":"Language acquisition","level":2,"score":0.25189998745918274}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2510.06826","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.06826","pdf_url":"https://arxiv.org/pdf/2510.06826","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2510.06826","doi":"https://doi.org/10.48550/arxiv.2510.06826","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.06826","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:2510.06826","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.06826","pdf_url":"https://arxiv.org/pdf/2510.06826","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","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/W4415317404.pdf","grobid_xml":"https://content.openalex.org/works/W4415317404.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7101794566","https://openalex.org/W7130510261","https://openalex.org/W7127569280","https://openalex.org/W7138953653","https://openalex.org/W7197050316","https://openalex.org/W4412569726","https://openalex.org/W4391555991","https://openalex.org/W7155680166","https://openalex.org/W4415336624","https://openalex.org/W4410371768"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2,25],"(LLMs)":[3],"are":[4],"typically":[5],"developed":[6],"through":[7,63],"large-scale":[8],"pre-training":[9],"followed":[10],"by":[11],"task-specific":[12],"fine-tuning.":[13],"Recent":[14],"advances":[15],"highlight":[16],"the":[17,67,99],"importance":[18],"of":[19,91,102],"an":[20],"intermediate":[21],"mid-training":[22,92,104],"stage,":[23],"where":[24],"undergo":[26],"multiple":[27],"annealing-style":[28],"phases":[29],"that":[30],"refine":[31],"data":[32,106],"quality,":[33],"adapt":[34],"optimization":[35],"schedules,":[36],"and":[37,51,70,77,110,120,134,140],"extend":[38],"context":[39],"length.":[40],"This":[41],"stage":[42],"mitigates":[43],"diminishing":[44],"returns":[45],"from":[46],"noisy":[47],"tokens,":[48],"stabilizes":[49],"convergence,":[50],"expands":[52],"model":[53],"capability":[54],"in":[55,82],"late":[56],"training.":[57],"Its":[58],"effectiveness":[59],"can":[60],"be":[61],"explained":[62],"gradient":[64],"noise":[65],"scale,":[66],"information":[68],"bottleneck,":[69],"curriculum":[71],"learning,":[72],"which":[73],"together":[74],"promote":[75],"generalization":[76],"abstraction.":[78],"Despite":[79],"widespread":[80],"use":[81],"state-of-the-art":[83],"systems,":[84],"there":[85],"has":[86],"been":[87],"no":[88],"prior":[89],"survey":[90],"as":[93],"a":[94],"unified":[95],"paradigm.":[96],"We":[97,113,129],"introduce":[98],"first":[100],"taxonomy":[101],"LLM":[103],"spanning":[105],"distribution,":[107],"learning-rate":[108],"scheduling,":[109],"long-context":[111],"extension.":[112],"distill":[114],"practical":[115],"insights,":[116],"compile":[117],"evaluation":[118],"benchmarks,":[119],"report":[121],"gains":[122],"to":[123],"enable":[124],"structured":[125],"comparisons":[126],"across":[127],"models.":[128],"also":[130],"identify":[131],"open":[132],"challenges":[133],"propose":[135],"avenues":[136],"for":[137],"future":[138],"research":[139],"practice.":[141]},"counts_by_year":[{"year":2026,"cited_by_count":7}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2025-10-18T00:00:00"}