{"id":"https://openalex.org/W7166802023","doi":"https://doi.org/10.18653/v1/2026.acl-long.799","title":"ETR: Entropy Trend Reward for Efficient Chain-of-Thought Reasoning","display_name":"ETR: Entropy Trend Reward for Efficient Chain-of-Thought Reasoning","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166802023","doi":"https://doi.org/10.18653/v1/2026.acl-long.799"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.acl-long.799","doi":"https://doi.org/10.18653/v1/2026.acl-long.799","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.799","pdf_url":"https://aclanthology.org/2026.acl-long.799.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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"},"volume":null,"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2026.acl-long.799.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133236603","display_name":"Xuan Xiong","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Xuan Xiong","raw_affiliation_strings":["University of Toronto"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Toronto","institution_ids":["https://openalex.org/I185261750"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139766905","display_name":"Huan Liu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98251732","display_name":"McMaster University","ror":"https://ror.org/02fa3aq29","country_code":"CA","type":"education","lineage":["https://openalex.org/I98251732"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Huan Liu","raw_affiliation_strings":["McMaster University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"McMaster University","institution_ids":["https://openalex.org/I98251732"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139797692","display_name":"Li Gu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I60158472","display_name":"Concordia University","ror":"https://ror.org/0420zvk78","country_code":"CA","type":"education","lineage":["https://openalex.org/I60158472"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Li Gu","raw_affiliation_strings":["Concordia University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Concordia University","institution_ids":["https://openalex.org/I60158472"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064393186","display_name":"Zhixiang Chi","orcid":"https://orcid.org/0000-0003-4560-4986","observed_orcids":["https://orcid.org/0000-0003-4560-4986"]},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Zhixiang Chi","raw_affiliation_strings":["University of Toronto"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Toronto","institution_ids":["https://openalex.org/I185261750"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139796497","display_name":"Yue Qiu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I153718931","display_name":"University of Ottawa","ror":"https://ror.org/03c4mmv16","country_code":"CA","type":"education","lineage":["https://openalex.org/I153718931"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Yue Qiu","raw_affiliation_strings":["University of Ottawa"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Ottawa","institution_ids":["https://openalex.org/I153718931"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139771579","display_name":"Yuanhao YU","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98251732","display_name":"McMaster University","ror":"https://ror.org/02fa3aq29","country_code":"CA","type":"education","lineage":["https://openalex.org/I98251732"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Yuanhao YU","raw_affiliation_strings":["McMaster University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"McMaster University","institution_ids":["https://openalex.org/I98251732"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5139842370","display_name":"Yang Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I60158472","display_name":"Concordia University","ror":"https://ror.org/0420zvk78","country_code":"CA","type":"education","lineage":["https://openalex.org/I60158472"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Yang Wang","raw_affiliation_strings":["Concordia University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Concordia University","institution_ids":["https://openalex.org/I60158472"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"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":0.43171393,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"17576","last_page":"17594"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.7529000043869019,"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":0.7529000043869019,"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":0.10989999771118164,"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":0.025699999183416367,"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/chain-of-thought-reasoning","display_name":"chain-of-thought reasoning","score":0.9909999966621399},{"id":"https://openalex.org/keywords/reasoning-efficiency","display_name":"reasoning efficiency","score":0.7979999780654907},{"id":"https://openalex.org/keywords/group-relative-policy-optimization","display_name":"Group Relative Policy Optimization","score":0.8999999761581421},{"id":"https://openalex.org/keywords/accuracy-efficiency-trade-off","display_name":"accuracy-efficiency trade-off","score":0.8769999742507935},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.7310000061988831}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4999000132083893},{"id":"https://openalex.org/C106301342","wikidata":"https://www.wikidata.org/wiki/Q4117933","display_name":"Entropy (arrow of time)","level":2,"score":0.4643999934196472},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4293999969959259},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3749000132083893},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2874999940395355},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.25369998812675476},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.24950000643730164},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.23849999904632568},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.23690000176429749},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2190999984741211}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.799","doi":"https://doi.org/10.18653/v1/2026.acl-long.799","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.799","pdf_url":"https://aclanthology.org/2026.acl-long.799.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.799","doi":"https://doi.org/10.18653/v1/2026.acl-long.799","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.799","pdf_url":"https://aclanthology.org/2026.acl-long.799.pdf","source":{"id":"https://openalex.org/S4393915740","display_name":"Proceedings of the conference - Association for Computational Linguistics. Meeting","issn_l":"0736-587X","issn":["0736-587X"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166802023.pdf","grobid_xml":"https://content.openalex.org/works/W7166802023.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7129076211","https://openalex.org/W7204945935","https://openalex.org/W7163046114","https://openalex.org/W4416027195","https://openalex.org/W7138946210","https://openalex.org/W7131869551","https://openalex.org/W7166798970","https://openalex.org/W7134844655","https://openalex.org/W7131126100","https://openalex.org/W4417299406"],"abstract_inverted_index":{"Chain-of-thought":[0],"(CoT)":[1],"reasoning":[2,18,42,94],"improves":[3],"large":[4],"language":[5],"model":[6],"performance":[7],"on":[8],"complex":[9],"tasks,":[10],"but":[11],"often":[12],"produces":[13],"excessively":[14],"long":[15],"and":[16,89,96],"inefficient":[17],"traces.Existing":[19],"methods":[20],"shorten":[21],"CoTs":[22],"using":[23],"length":[24,113],"penalties":[25],"or":[26],"global":[27],"entropy":[28,54],"reduction,":[29],"implicitly":[30],"assuming":[31],"that":[32,41,71],"low":[33],"uncertainty":[34,74],"is":[35,44],"desirable":[36],"throughout":[37],"reasoning.We":[38],"show":[39],"instead":[40],"efficiency":[43],"governed":[45],"by":[46,59,107,114],"the":[47],"trajectory":[48],"of":[49],"uncertainty.CoTs":[50],"with":[51],"dominant":[52],"downward":[53],"trends":[55],"are":[56],"substantially":[57],"shorter.Motivated":[58],"this":[60],"insight,":[61],"we":[62],"propose":[63],"Entropy":[64],"Trend":[65],"Reward":[66],"(ETR),":[67],"a":[68,101],"trajectory-aware":[69],"objective":[70],"encourages":[72],"progressive":[73],"reduction":[75],"while":[76,110],"allowing":[77],"limited":[78],"local":[79],"exploration.We":[80],"integrate":[81],"ETR":[82],"into":[83],"Group":[84],"Relative":[85],"Policy":[86],"Optimization":[87],"(GRPO)":[88],"evaluate":[90],"it":[91],"across":[92,116],"multiple":[93],"models":[95],"challenging":[97],"benchmarks.ETR":[98],"consistently":[99],"achieves":[100],"superior":[102],"accuracy-efficiency":[103],"trade-off,":[104],"improving":[105],"DeepSeek-R1-Distill-7B":[106],"+9.9%":[108],"accuracy":[109],"reducing":[111],"CoT":[112],"67%":[115],"four":[117],"benchmarks.":[118]},"counts_by_year":[],"updated_date":"2026-10-29T09:56:32.521200","created_date":"2026-07-02T00:00:00"}