{"id":"https://openalex.org/W4412889951","doi":"https://doi.org/10.18653/v1/2025.acl-long.1278","title":"Balancing Diversity and Risk in LLM Sampling: How to Select Your Method and Parameter for Open-Ended Text Generation","display_name":"Balancing Diversity and Risk in LLM Sampling: How to Select Your Method and Parameter for Open-Ended Text Generation","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412889951","doi":"https://doi.org/10.18653/v1/2025.acl-long.1278"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-long.1278","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1278","pdf_url":"https://aclanthology.org/2025.acl-long.1278.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.1278.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5025994913","display_name":"Yuxuan Zhou","orcid":"https://orcid.org/0000-0001-5091-4431"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuxuan Zhou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029656834","display_name":"Margret Keuper","orcid":"https://orcid.org/0000-0002-8437-7993"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Margret Keuper","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5003887059","display_name":"Mario Fritz","orcid":"https://orcid.org/0000-0001-8949-9896"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mario Fritz","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":4,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"26352","last_page":"26365"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9300000071525574,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9300000071525574,"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":0.6729196906089783},{"id":"https://openalex.org/keywords/diversity","display_name":"Diversity (politics)","score":0.5842391848564148},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.4403689205646515},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.12987646460533142},{"id":"https://openalex.org/keywords/political-science","display_name":"Political science","score":0.07454770803451538}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6729196906089783},{"id":"https://openalex.org/C2781316041","wikidata":"https://www.wikidata.org/wiki/Q1230584","display_name":"Diversity (politics)","level":2,"score":0.5842391848564148},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.4403689205646515},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.12987646460533142},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.07454770803451538},{"id":"https://openalex.org/C94915269","wikidata":"https://www.wikidata.org/wiki/Q1834857","display_name":"Detector","level":2,"score":0.0},{"id":"https://openalex.org/C199539241","wikidata":"https://www.wikidata.org/wiki/Q7748","display_name":"Law","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.acl-long.1278","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1278","pdf_url":"https://aclanthology.org/2025.acl-long.1278.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.1278","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1278","pdf_url":"https://aclanthology.org/2025.acl-long.1278.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":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412889951.pdf","grobid_xml":"https://content.openalex.org/works/W4412889951.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":{"Sampling-based":[0],"decoding":[1,104],"strategies":[2],"have":[3,53],"been":[4,54],"widely":[5],"adopted":[6],"for":[7,137],"Large":[8],"Language":[9],"Models":[10],"(LLMs)":[11],"in":[12],"numerous":[13],"applications,":[14],"targeting":[15],"a":[16,82,90,117,122,133],"balance":[17],"between":[18,98],"diversity":[19,99],"and":[20,25,73,100,130],"quality":[21],"via":[22],"temperature":[23],"tuning":[24],"tail":[26,46],"truncation.Considering":[27],"the":[28,32,45,64,70,74,87,96,114],"strong":[29],"dependency":[30],"of":[31,47,89,116,125],"candidate":[33],"next":[34],"tokens":[35],"on":[36,59,69,107],"different":[37],"prefixes,":[38],"recent":[39],"studies":[40],"propose":[41,81],"to":[42,85],"adaptively":[43],"truncate":[44],"LLMs'":[48],"predicted":[49],"distribution.Although":[50],"improved":[51],"results":[52,65],"reported":[55],"with":[56],"these":[57],"methods":[58,129],"open-ended":[60],"text":[61],"generation":[62],"tasks,":[63],"are":[66],"highly":[67],"dependent":[68],"curated":[71],"parameters":[72],"limited":[75],"exemplar":[76],"text.In":[77],"this":[78],"paper,":[79],"we":[80],"systematic":[83],"way":[84],"estimate":[86],"capacity":[88],"truncation":[91,127],"sampling":[92,128],"method":[93],"by":[94],"considering":[95],"trade-off":[97],"risk":[101],"at":[102,144],"each":[103],"step,":[105],"based":[106],"our":[108],"collected":[109],"prefix":[110],"tree":[111],"which":[112],"preserves":[113],"context":[115],"full":[118],"sentence.Our":[119],"work":[120],"offers":[121],"comprehensive":[123],"comparison":[124],"existing":[126],"serves":[131],"as":[132],"practical":[134],"user":[135],"guideline":[136],"their":[138],"parameter":[139],"selection.Our":[140],"code":[141],"is":[142],"available":[143],"github":[145],"repository.":[146]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}