{"id":"https://openalex.org/W4415036737","doi":"https://doi.org/10.48550/arxiv.2505.21067","title":"Why Distillation can Outperform Zero-RL: The Role of Flexible Reasoning","display_name":"Why Distillation can Outperform Zero-RL: The Role of Flexible Reasoning","publication_year":2025,"publication_date":"2025-05-27","ids":{"openalex":"https://openalex.org/W4415036737","doi":"https://doi.org/10.48550/arxiv.2505.21067"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2505.21067","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2505.21067","pdf_url":"https://arxiv.org/pdf/2505.21067","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"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2505.21067","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5085425510","display_name":"Xiaojun Hu","orcid":"https://orcid.org/0000-0001-7092-2535","observed_orcids":["https://orcid.org/0000-0001-7092-2535"]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hu, Xiao","raw_affiliation_strings":["Tsinghua University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University ,","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064644585","display_name":"Xingyu Lu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lu, Xingyu","raw_affiliation_strings":["Tsinghua University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University ,","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071122369","display_name":"Liyuan Mao","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Mao, Liyuan","raw_affiliation_strings":["Shanghai Jiao Tong University , 3 CASIA ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University , 3 CASIA ,","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhang, YiFan","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, YiFan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008792727","display_name":"Tianke Zhang","orcid":"https://orcid.org/0009-0003-1270-200X","observed_orcids":["https://orcid.org/0009-0003-1270-200X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Tianke","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101739645","display_name":"Bin Wen","orcid":"https://orcid.org/0000-0003-3449-5841","observed_orcids":["https://orcid.org/0000-0003-3449-5841"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wen, Bin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5067183063","display_name":"Fan Yang","orcid":"https://orcid.org/0000-0002-3431-2585","observed_orcids":["https://orcid.org/0000-0002-3431-2585","https://orcid.org/0009-0003-7030-4309"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Fan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112381636","display_name":"Tingting Gao","orcid":"https://orcid.org/0000-0003-2765-4963","observed_orcids":["https://orcid.org/0000-0003-2765-4963"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Tingting","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5060887535","display_name":"G.J. Zhou","orcid":"https://orcid.org/0000-0002-0277-4638","observed_orcids":["https://orcid.org/0000-0002-0277-4638"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Guorui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":17,"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":0.8457000255584717,"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":0.8457000255584717,"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":0.065700002014637,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.01590000092983246,"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/knowledge-distillation","display_name":"knowledge distillation","score":0.8510000109672546},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.9539999961853027},{"id":"https://openalex.org/keywords/multi-perspective-thinking","display_name":"multi-perspective thinking","score":0.8769999742507935},{"id":"https://openalex.org/keywords/metacognitive-awareness","display_name":"metacognitive awareness","score":0.9959999918937683},{"id":"https://openalex.org/keywords/logical-connectives","display_name":"logical connectives","score":0.6480000019073486}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.729200005531311},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.6862000226974487},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5989000201225281},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.498199999332428},{"id":"https://openalex.org/C169900460","wikidata":"https://www.wikidata.org/wiki/Q2200417","display_name":"Cognition","level":2,"score":0.453900009393692},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.42800000309944153},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.4196999967098236},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41589999198913574},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.41519999504089355},{"id":"https://openalex.org/C43971567","wikidata":"https://www.wikidata.org/wiki/Q3142865","display_name":"Logical reasoning","level":2,"score":0.37869998812675476},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.3763999938964844},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2921999990940094},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.274399995803833},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.2720000147819519},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.2630000114440918},{"id":"https://openalex.org/C161407221","wikidata":"https://www.wikidata.org/wiki/Q4382939","display_name":"Cognitive model","level":3,"score":0.2606000006198883}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2505.21067","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2505.21067","pdf_url":"https://arxiv.org/pdf/2505.21067","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.2505.21067","doi":"https://doi.org/10.48550/arxiv.2505.21067","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2505.21067","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:2505.21067","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2505.21067","pdf_url":"https://arxiv.org/pdf/2505.21067","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":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4415036737.pdf","grobid_xml":"https://content.openalex.org/works/W4415036737.grobid-xml"},"referenced_works_count":27,"referenced_works":["https://openalex.org/W2736601468","https://openalex.org/W3083410900","https://openalex.org/W3134642945","https://openalex.org/W3177813494","https://openalex.org/W4223908421","https://openalex.org/W4391631327","https://openalex.org/W4400141325","https://openalex.org/W4403746689","https://openalex.org/W4404330874","https://openalex.org/W4405655184","https://openalex.org/W4405766390","https://openalex.org/W4405903187","https://openalex.org/W4406735009","https://openalex.org/W4406774573","https://openalex.org/W4407231762","https://openalex.org/W4407719847","https://openalex.org/W4409049124","https://openalex.org/W4414281281","https://openalex.org/W4414973451","https://openalex.org/W4415976483","https://openalex.org/W4416036482","https://openalex.org/W4416118517","https://openalex.org/W4417055076","https://openalex.org/W6929276109","https://openalex.org/W6948343298","https://openalex.org/W7196953391","https://openalex.org/W7196963086"],"related_works":["https://openalex.org/W7166800326","https://openalex.org/W7138915204","https://openalex.org/W4415022868","https://openalex.org/W6966810623","https://openalex.org/W7160660473","https://openalex.org/W7131853216","https://openalex.org/W7119997872","https://openalex.org/W4417536778","https://openalex.org/W6948343298","https://openalex.org/W7135052881"],"abstract_inverted_index":{"Reinforcement":[0],"learning":[1],"(RL)":[2],"has":[3],"played":[4],"an":[5],"important":[6],"role":[7],"in":[8,36,73],"improving":[9],"the":[10,52,70,79,97,106,147],"reasoning":[11,139],"ability":[12],"of":[13,108,122,149],"large":[14],"language":[15],"models":[16,26],"(LLMs).":[17],"Some":[18],"studies":[19],"apply":[20],"RL":[21],"directly":[22],"to":[23,130,144],"\\textit{smaller}":[24],"base":[25,53],"(known":[27],"as":[28],"zero-RL)":[29],"and":[30,65,90,117],"also":[31],"achieve":[32],"notable":[33],"progress.":[34],"However,":[35],"this":[37],"paper,":[38],"we":[39,76],"show":[40],"that":[41,78,103],"using":[42],"only":[43],"920":[44],"examples,":[45],"a":[46],"simple":[47],"distillation":[48,104],"method":[49],"based":[50],"on":[51],"model":[54,74,81],"can":[55],"clearly":[56],"outperform":[57],"zero-RL,":[58],"which":[59,133],"typically":[60],"requires":[61],"much":[62,93],"more":[63,83,94],"data":[64],"computational":[66],"cost.":[67],"By":[68],"analyzing":[69],"token":[71],"frequency":[72,148],"outputs,":[75],"find":[77],"distilled":[80],"shows":[82],"flexible":[84,131],"reasoning.":[85],"It":[86],"uses":[87],"anthropomorphic":[88],"tokens":[89],"logical":[91],"connectors":[92],"often":[95],"than":[96],"zero-RL":[98,142],"model.":[99],"Further":[100],"analysis":[101],"reveals":[102],"enhances":[105],"presence":[107],"two":[109,124],"advanced":[110,125],"cognitive":[111,126],"behaviors:":[112],"Multi-Perspective":[113],"Thinking":[114],"or":[115],"Attempting":[116],"Metacognitive":[118],"Awareness.":[119],"Frequent":[120],"occurrences":[121],"these":[123,150],"behaviors":[127],"give":[128],"rise":[129],"reasoning,":[132],"is":[134],"essential":[135],"for":[136],"solving":[137],"complex":[138],"problems,":[140],"while":[141],"fails":[143],"significantly":[145],"boost":[146],"behaviors.":[151]},"counts_by_year":[{"year":2026,"cited_by_count":10},{"year":2025,"cited_by_count":7}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-10T00:00:00"}