{"id":"https://openalex.org/W4412944734","doi":"https://doi.org/10.18653/v1/2025.acl-long.1137","title":"SoftCoT: Soft Chain-of-Thought for Efficient Reasoning with LLMs","display_name":"SoftCoT: Soft Chain-of-Thought for Efficient Reasoning with LLMs","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412944734","doi":"https://doi.org/10.18653/v1/2025.acl-long.1137"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-long.1137","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1137","pdf_url":"https://aclanthology.org/2025.acl-long.1137.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 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.acl-long.1137.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5055389051","display_name":"Yaogeng Xu","orcid":"https://orcid.org/0009-0009-5025-8118"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yige Xu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104089978","display_name":"Xu Guo","orcid":"https://orcid.org/0009-0004-1258-8232"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu Guo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5116335603","display_name":"Zhiwei Zeng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhiwei Zeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5116335604","display_name":"Chunyan Miao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chunyan Miao","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":9,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"23336","last_page":"23351"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10215","display_name":"Semantic Web and Ontologies","score":0.9460999965667725,"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/T10215","display_name":"Semantic Web and Ontologies","score":0.9460999965667725,"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/T10317","display_name":"Advanced Database Systems and Queries","score":0.9172000288963318,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T11010","display_name":"Logic, Reasoning, and Knowledge","score":0.9156000018119812,"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/computer-science","display_name":"Computer science","score":0.5514460802078247},{"id":"https://openalex.org/keywords/chain","display_name":"Chain (unit)","score":0.4807038903236389}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5514460802078247},{"id":"https://openalex.org/C199185054","wikidata":"https://www.wikidata.org/wiki/Q552299","display_name":"Chain (unit)","level":2,"score":0.4807038903236389},{"id":"https://openalex.org/C1276947","wikidata":"https://www.wikidata.org/wiki/Q333","display_name":"Astronomy","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.acl-long.1137","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1137","pdf_url":"https://aclanthology.org/2025.acl-long.1137.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 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.acl-long.1137","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1137","pdf_url":"https://aclanthology.org/2025.acl-long.1137.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 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G3637000134","display_name":null,"funder_award_id":"NRF-NRFI05-2019-0002","funder_id":"https://openalex.org/F4320320709","funder_display_name":"National Research Foundation Singapore"}],"funders":[{"id":"https://openalex.org/F4320320671","display_name":"National Research Foundation","ror":"https://ror.org/05s0g1g46"},{"id":"https://openalex.org/F4320320709","display_name":"National Research Foundation Singapore","ror":"https://ror.org/03cpyc314"},{"id":"https://openalex.org/F4320320766","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412944734.pdf","grobid_xml":"https://content.openalex.org/works/W4412944734.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Chain-of-Thought":[0],"(CoT)":[1],"reasoning":[2,11,16,28,82,127,135],"enables":[3],"Large":[4],"Language":[5],"Models":[6],"(LLMs)":[7],"to":[8,58,97],"solve":[9],"complex":[10],"tasks":[12],"by":[13],"generating":[14],"intermediate":[15],"steps.However,":[17],"most":[18],"existing":[19],"approaches":[20],"focus":[21],"on":[22,125],"hard":[23],"token":[24],"decoding,":[25],"which":[26,110],"constrains":[27],"within":[29],"the":[30,88,105,115],"discrete":[31],"vocabulary":[32],"space":[33,118],"and":[34,50],"may":[35],"not":[36,85],"always":[37],"be":[38],"optimal.While":[39],"recent":[40],"efforts":[41],"explore":[42],"continuousspace":[43],"reasoning,":[44],"they":[45],"often":[46],"require":[47,86],"full-model":[48],"fine-tuning":[49],"suffer":[51],"from":[52],"catastrophic":[53],"forgetting,":[54],"limiting":[55],"their":[56],"applicability":[57],"state-of-theart":[59],"LLMs":[60],"that":[61,83,130],"already":[62],"perform":[63],"well":[64],"in":[65],"zeroshot":[66],"settings":[67],"with":[68],"a":[69,77,92,120],"proper":[70],"instruction.To":[71],"address":[72],"this":[73],"challenge,":[74],"we":[75,90],"propose":[76],"novel":[78],"approach":[79],"for":[80],"continuous-space":[81],"does":[84],"modifying":[87],"LLM.Specifically,":[89],"employ":[91],"lightweight":[93],"fixed":[94],"assistant":[95],"model":[96],"speculatively":[98],"generate":[99],"instancespecific":[100],"soft":[101],"thought":[102],"tokens":[103],"as":[104],"initial":[106],"chain":[107],"of":[108],"thoughts,":[109],"are":[111],"then":[112],"mapped":[113],"into":[114],"LLM's":[116],"representation":[117],"via":[119],"trainable":[121],"projection":[122],"module.Experimental":[123],"results":[124],"five":[126],"benchmarks":[128],"demonstrate":[129],"our":[131],"method":[132],"enhances":[133],"LLM":[134],"performance":[136],"through":[137],"supervised,":[138],"parameter-efficient":[139],"fine-tuning.":[140]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":7}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}