{"id":"https://openalex.org/W4416400869","doi":"https://doi.org/10.18653/v1/2026.acl-long.122","title":"LightReasoner: Can Small Language Models Teach Large Language Models Reasoning?","display_name":"LightReasoner: Can Small Language Models Teach Large Language Models Reasoning?","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W4416400869","doi":"https://doi.org/10.18653/v1/2026.acl-long.122"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.acl-long.122","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.122","pdf_url":"https://aclanthology.org/2026.acl-long.122.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.acl-long.122.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100400839","display_name":"Jingyuan Wang","orcid":"https://orcid.org/0000-0002-0903-1896"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jingyuan Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5053712210","display_name":"Yankai Chen","orcid":"https://orcid.org/0000-0001-5741-2047"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yankai Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023084303","display_name":"Zhonghang Li","orcid":"https://orcid.org/0000-0002-3977-1334"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhonghang Li","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5073902371","display_name":"Chao Huang","orcid":"https://orcid.org/0000-0002-1521-5739"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chao Huang","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.01012146,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"2635","last_page":"2663"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.17560000717639923,"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"}},"topics":[{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.17560000717639923,"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/T10028","display_name":"Topic Modeling","score":0.1535000056028366,"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/T13910","display_name":"Computational and Text Analysis Methods","score":0.09669999778270721,"subfield":{"id":"https://openalex.org/subfields/3300","display_name":"General Social Sciences"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.6945000290870667},{"id":"https://openalex.org/keywords/counterintuitive","display_name":"Counterintuitive","score":0.6004999876022339},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.444599986076355},{"id":"https://openalex.org/keywords/phrase","display_name":"Phrase","score":0.4221000075340271},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.41830000281333923},{"id":"https://openalex.org/keywords/fraction","display_name":"Fraction (chemistry)","score":0.38089999556541443},{"id":"https://openalex.org/keywords/paraphrase","display_name":"Paraphrase","score":0.3531999886035919},{"id":"https://openalex.org/keywords/measure","display_name":"Measure (data warehouse)","score":0.32679998874664307},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.3228999972343445}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7703999876976013},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6945000290870667},{"id":"https://openalex.org/C101097943","wikidata":"https://www.wikidata.org/wiki/Q5176983","display_name":"Counterintuitive","level":2,"score":0.6004999876022339},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5432999730110168},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.444599986076355},{"id":"https://openalex.org/C2776224158","wikidata":"https://www.wikidata.org/wiki/Q187931","display_name":"Phrase","level":2,"score":0.4221000075340271},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.41830000281333923},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3968000113964081},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.38269999623298645},{"id":"https://openalex.org/C149629883","wikidata":"https://www.wikidata.org/wiki/Q660926","display_name":"Fraction (chemistry)","level":2,"score":0.38089999556541443},{"id":"https://openalex.org/C2780922921","wikidata":"https://www.wikidata.org/wiki/Q255189","display_name":"Paraphrase","level":2,"score":0.3531999886035919},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.32679998874664307},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.3228999972343445},{"id":"https://openalex.org/C207390915","wikidata":"https://www.wikidata.org/wiki/Q1230525","display_name":"Divergence (linguistics)","level":2,"score":0.3188999891281128},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.30300000309944153},{"id":"https://openalex.org/C129792486","wikidata":"https://www.wikidata.org/wiki/Q1050419","display_name":"Language identification","level":3,"score":0.2980000078678131},{"id":"https://openalex.org/C2778334786","wikidata":"https://www.wikidata.org/wiki/Q1586270","display_name":"Variation (astronomy)","level":2,"score":0.29019999504089355},{"id":"https://openalex.org/C74672266","wikidata":"https://www.wikidata.org/wiki/Q815859","display_name":"Language acquisition","level":2,"score":0.2840999960899353},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.2840000092983246},{"id":"https://openalex.org/C2983448237","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Language understanding","level":2,"score":0.28200000524520874},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.2653999924659729},{"id":"https://openalex.org/C155092808","wikidata":"https://www.wikidata.org/wiki/Q182557","display_name":"Computational linguistics","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.25380000472068787},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.25279998779296875},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.25060001015663147}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.122","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.122","pdf_url":"https://aclanthology.org/2026.acl-long.122.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2510.07962","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2510.07962","pdf_url":"https://arxiv.org/pdf/2510.07962","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,"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.2510.07962","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.07962","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,"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:10.18653/v1/2026.acl-long.122","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.122","pdf_url":"https://aclanthology.org/2026.acl-long.122.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416400869.pdf","grobid_xml":"https://content.openalex.org/works/W4416400869.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1,48,53],"models":[2,49,54],"(LLMs)":[3,55],"have":[4],"demonstrated":[5],"remarkable":[6],"progress":[7],"in":[8,92],"reasoning,":[9],"often":[10],"through":[11,112],"supervised":[12],"fine-tuning":[13,118],"(SFT).However,":[14],"SFT":[15],"is":[16],"resource-intensive,":[17],"relying":[18,160],"on":[19,161],"large":[20],"curated":[21],"datasets,":[22],"rejection-sampled":[23],"demonstrations,":[24],"and":[25,85,104,115,152,175],"uniform":[26],"optimization":[27],"across":[28],"all":[29,158],"tokens-even":[30],"though":[31],"only":[32],"a":[33,43,71,80,86,96,117,173],"fraction":[34],"carry":[35],"meaningful":[36],"learning":[37],"value.In":[38],"this":[39],"work,":[40],"we":[41],"explore":[42],"counterintuitive":[44],"idea:":[45],"can":[46],"smaller":[47],"(SLMs)":[50],"teach":[51],"larger":[52],"by":[56,138,146,150,156],"revealing":[57],"high-value":[58],"reasoning":[59,102,131],"moments":[60,103],"that":[61,74,99,120],"reflect":[62],"the":[63,76,109,122],"latter's":[64],"unique":[65],"strength?We":[66],"propose":[67],"LightReasoner":[68,135,171],"1":[69],",":[70],"novel":[72],"framework":[73],"leverages":[75],"behavioral":[77],"divergence":[78],"between":[79],"stronger":[81],"expert":[82,123],"model":[83,89,124],"(LLM)":[84],"weaker":[87,165],"amateur":[88],"(SLM).LightReasoner":[90],"operates":[91],"two":[93],"stages:":[94],"(1)":[95],"sampling":[97],"stage":[98,119],"pinpoints":[100],"critical":[101],"constructs":[105],"supervision":[106],"examples":[107],"capturing":[108],"expert's":[110],"advantage":[111],"expert-amateur":[113],"contrast,":[114],"(2)":[116],"aligns":[121],"with":[125],"these":[126],"distilled":[127],"examples,":[128],"amplifying":[129],"its":[130],"strengths.Across":[132],"seven":[133],"benchmarks,":[134],"improves":[136],"accuracy":[137],"up":[139],"to":[140],"28.1%,":[141],"while":[142],"reducing":[143],"time":[144],"consumption":[145],"90%,":[147],"sampled":[148],"problems":[149],"80%,":[151],"tuned":[153],"token":[154],"usage":[155],"99%,":[157],"without":[159],"ground-truth":[162],"labels.By":[163],"turning":[164],"SLMs":[166],"into":[167],"effective":[168],"teaching":[169],"signals,":[170],"offers":[172],"scalable":[174],"resource-efficient":[176],"approach":[177],"for":[178],"advancing":[179],"LLM":[180],"reasoning.":[181]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-11T00:00:00"}