{"id":"https://openalex.org/W4415799149","doi":"https://doi.org/10.52202/079017-0059","title":"JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models","display_name":"JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models","publication_year":2024,"publication_date":"2024-01-01","ids":{"openalex":"https://openalex.org/W4415799149","doi":"https://doi.org/10.52202/079017-0059"},"language":"en","primary_location":{"id":"doi:10.52202/079017-0059","doi":"https://doi.org/10.52202/079017-0059","is_oa":false,"landing_page_url":"https://doi.org/10.52202/079017-0059","pdf_url":null,"source":{"id":"https://openalex.org/S4363606243","display_name":"neural information processing systems","issn_l":"1049-5258","issn":["1049-5258"],"is_oa":false,"is_in_doaj":false,"is_core":true,"listed_in":["cwts-core","norway-2"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Advances in Neural Information Processing Systems 37","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2405.14365","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5063459528","display_name":"Kun Zhou","orcid":"https://orcid.org/0000-0003-0650-9521","observed_orcids":["https://orcid.org/0000-0003-0650-9521"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kun Zhou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073554794","display_name":"Beichen Zhang","orcid":"https://orcid.org/0000-0001-8093-3887","observed_orcids":["https://orcid.org/0000-0001-8093-3887"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Beichen Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5115695120","display_name":"Jiapeng Wang","orcid":"https://orcid.org/0000-0003-1498-0305","observed_orcids":["https://orcid.org/0000-0003-1498-0305"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiapeng Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100325084","display_name":"Zhipeng Chen","orcid":"https://orcid.org/0000-0002-8330-0070","observed_orcids":["https://orcid.org/0000-0002-8330-0070"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhipeng Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037145565","display_name":"Wayne Xin Zhao","orcid":"https://orcid.org/0000-0002-8333-6196","observed_orcids":["https://orcid.org/0000-0002-8333-6196"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wayne Zhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027741558","display_name":"Jing Jie Sha","orcid":"https://orcid.org/0009-0007-5295-3857","observed_orcids":["https://orcid.org/0009-0007-5295-3857","https://orcid.org/0009-0000-5623-3549"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jing Sha","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zhichao Sheng","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhichao Sheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103143080","display_name":"Shijin Wang","orcid":"https://orcid.org/0000-0002-4340-7420","observed_orcids":["https://orcid.org/0000-0002-4340-7420","https://orcid.org/0009-0009-2562-5721"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shijin Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5025631695","display_name":"Ji-Rong Wen","orcid":"https://orcid.org/0000-0002-9777-9676","observed_orcids":["https://orcid.org/0000-0002-9777-9676"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji-Rong Wen","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":1.4436,"has_fulltext":true,"cited_by_count":5,"citation_normalized_percentile":{"value":0.85358449,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"1854","last_page":"1889"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.1956000030040741,"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":0.1956000030040741,"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/T13523","display_name":"Mathematics, Computing, and Information Processing","score":0.14110000431537628,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T12522","display_name":"Mathematics Education and Pedagogy","score":0.12600000202655792,"subfield":{"id":"https://openalex.org/subfields/2604","display_name":"Applied Mathematics"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/mathematical-reasoning","display_name":"mathematical reasoning","score":0.8360000252723694},{"id":"https://openalex.org/keywords/synthetic-data-generation","display_name":"synthetic data generation","score":0.5989999771118164},{"id":"https://openalex.org/keywords/training-efficiency","display_name":"training efficiency","score":0.4650000035762787}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6707000136375427},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5181000232696533},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.47920000553131104},{"id":"https://openalex.org/C2779280203","wikidata":"https://www.wikidata.org/wiki/Q17121211","display_name":"Small data","level":2,"score":0.4058000147342682},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.37279999256134033},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36230000853538513},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.2992999851703644},{"id":"https://openalex.org/C76969082","wikidata":"https://www.wikidata.org/wiki/Q486902","display_name":"Mathematical model","level":2,"score":0.2922999858856201},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.2799000144004822},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2784999907016754},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.27619999647140503},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.2606000006198883}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.52202/079017-0059","doi":"https://doi.org/10.52202/079017-0059","is_oa":false,"landing_page_url":"https://doi.org/10.52202/079017-0059","pdf_url":null,"source":{"id":"https://openalex.org/S4363606243","display_name":"neural information processing systems","issn_l":"1049-5258","issn":["1049-5258"],"is_oa":false,"is_in_doaj":false,"is_core":true,"listed_in":["cwts-core","norway-2"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Advances in Neural Information Processing Systems 37","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2405.14365","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2405.14365","pdf_url":"https://arxiv.org/pdf/2405.14365","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.2405.14365","doi":"https://doi.org/10.48550/arxiv.2405.14365","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2405.14365","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:2405.14365","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2405.14365","pdf_url":"https://arxiv.org/pdf/2405.14365","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":[{"id":"https://metadata.un.org/sdg/4","score":0.5385000109672546,"display_name":"Quality education"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4415799149.pdf","grobid_xml":"https://content.openalex.org/works/W4415799149.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4403578767","https://openalex.org/W4392222438","https://openalex.org/W7140132769","https://openalex.org/W4403928588","https://openalex.org/W7078545102","https://openalex.org/W4409348209","https://openalex.org/W4412945228","https://openalex.org/W4402670720","https://openalex.org/W7166817036","https://openalex.org/W4392538380"],"abstract_inverted_index":{"Mathematical":[0],"reasoning":[1,197,203],"is":[2],"an":[3,62],"important":[4],"capability":[5,95],"of":[6,40,105],"large":[7,45],"language":[8,202],"models~(LLMs)":[9],"for":[10,24,70,146,166],"real-world":[11],"applications.":[12],"To":[13,51,81],"enhance":[14],"this":[15],"capability,":[16],"existing":[17],"work":[18,41],"either":[19],"collects":[20],"large-scale":[21],"math-related":[22,138],"texts":[23],"pre-training,":[25],"or":[26,49],"relies":[27],"on":[28,56,108,182,194],"stronger":[29],"LLMs":[30],"(\\eg":[31],"GPT-4)":[32],"to":[33,44,74,90,112,115,133,152,160,174],"synthesize":[34,116,161],"massive":[35],"math":[36,71,120,164],"problems.":[37],"Both":[38],"types":[39],"generally":[42],"lead":[43],"costs":[46],"in":[47,216],"training":[48],"synthesis.":[50],"reduce":[52],"the":[53,97,128,135,148,154],"cost,":[54],"based":[55,107],"open-source":[57],"available":[58],"texts,":[59],"we":[60,84,101,126],"propose":[61],"efficient":[63],"way":[64],"that":[65,189],"trains":[66],"a":[67,86,103],"small":[68,98,155],"LLM":[69],"problem":[72],"synthesis,":[73],"efficiently":[75],"generate":[76],"sufficient":[77],"high-quality":[78],"pre-training":[79,167],"data.":[80,184],"achieve":[82],"it,":[83],"create":[85],"dataset":[87,151],"using":[88],"GPT-4":[89,145,176],"distill":[91],"its":[92],"data":[93,211],"synthesis":[94],"into":[96,144],"LLM.":[99,156],"Concretely,":[100],"craft":[102],"set":[104],"prompts":[106],"human":[109],"education":[110],"stages":[111],"guide":[113],"GPT-4,":[114],"problems":[117,165],"covering":[118],"diverse":[119],"knowledge":[121,149],"and":[122,180,204,210],"difficulty":[123],"levels.":[124],"Besides,":[125],"adopt":[127],"gradient-based":[129],"influence":[130],"estimation":[131],"method":[132],"select":[134],"most":[136],"valuable":[137],"texts.":[139],"The":[140],"both":[141,200],"are":[142],"fed":[143],"creating":[147],"distillation":[150],"train":[153],"We":[157],"leverage":[158],"it":[159],"6":[162],"million":[163],"our":[168],"JiuZhang3.0":[169,190],"model,":[170],"which":[171],"only":[172],"needs":[173],"invoke":[175],"API":[177],"9.3k":[178],"times":[179],"pre-train":[181],"4.6B":[183],"Experimental":[185],"results":[186],"have":[187],"shown":[188],"achieves":[191],"state-of-the-art":[192],"performance":[193],"several":[195],"mathematical":[196],"datasets,":[198],"under":[199],"natural":[201],"tool":[205],"manipulation":[206],"settings.":[207],"Our":[208],"code":[209],"will":[212],"be":[213],"publicly":[214],"released":[215],"\\url{https://github.com/RUCAIBox/JiuZhang3.0}.":[217]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":2}],"updated_date":"2026-10-13T10:08:12.033832","created_date":"2025-11-03T00:00:00"}