{"id":"https://openalex.org/W4412886958","doi":"https://doi.org/10.18653/v1/2025.acl-industry.2","title":"Speculative Reward Model Boosts Decision Making Ability of LLMs Cost-Effectively","display_name":"Speculative Reward Model Boosts Decision Making Ability of LLMs Cost-Effectively","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412886958","doi":"https://doi.org/10.18653/v1/2025.acl-industry.2"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-industry.2","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-industry.2","pdf_url":"https://aclanthology.org/2025.acl-industry.2.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 6: Industry Track)","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-industry.2.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5072908486","display_name":"Jing Gu","orcid":"https://orcid.org/0000-0002-0291-3263"},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiawei Gu","raw_affiliation_strings":["sen University Sun Yat-sen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"sen University Sun Yat-sen University","institution_ids":["https://openalex.org/I157773358"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5108952582","display_name":"Shangsong Liang","orcid":null},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shangsong Liang","raw_affiliation_strings":["sen University Sun Yat-sen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"sen University Sun Yat-sen University","institution_ids":["https://openalex.org/I157773358"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I157773358"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.16458692,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"4","last_page":"21"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13643","display_name":"Artificial Intelligence in Law","score":0.8215000033378601,"subfield":{"id":"https://openalex.org/subfields/3320","display_name":"Political Science and International Relations"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T13643","display_name":"Artificial Intelligence in Law","score":0.8215000033378601,"subfield":{"id":"https://openalex.org/subfields/3320","display_name":"Political Science and International Relations"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T13999","display_name":"Digital Rights Management and Security","score":0.7013000249862671,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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":0.46004724502563477},{"id":"https://openalex.org/keywords/risk-analysis","display_name":"Risk analysis (engineering)","score":0.3705652356147766},{"id":"https://openalex.org/keywords/business","display_name":"Business","score":0.32291287183761597}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.46004724502563477},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.3705652356147766},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.32291287183761597}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.acl-industry.2","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-industry.2","pdf_url":"https://aclanthology.org/2025.acl-industry.2.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 6: Industry Track)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.acl-industry.2","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-industry.2","pdf_url":"https://aclanthology.org/2025.acl-industry.2.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 6: Industry Track)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.5299999713897705,"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412886958.pdf","grobid_xml":"https://content.openalex.org/works/W4412886958.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":{"Effective":[0],"decision-making":[1,57,117],"in":[2,126,162,164],"Large":[3],"Language":[4],"Models":[5],"(LLMs)":[6],"is":[7,98,204,230],"essential":[8],"for":[9,51,157],"handling":[10],"intricate":[11],"tasks.However,":[12],"existing":[13,45,75],"approaches":[14],"prioritize":[15],"performance":[16,53],"but":[17],"often":[18,47],"overlook":[19],"the":[20,32,38,63,106,138,167,170,194,201,209,212,216,221,227],"balance":[21],"between":[22],"effectiveness":[23],"and":[24,104,123,159],"computational":[25],"cost.To":[26],"address":[27],"this,":[28],"we":[29,61],"first":[30],"introduce":[31],"3E":[33],"Criteria":[34],"to":[35,84,100,135,149],"systematically":[36],"assess":[37],"costeffectiveness":[39],"of":[40,137,169,211,220],"search":[41,76,107,140],"strategies,":[42],"revealing":[43],"that":[44,71,131,192],"methods":[46],"trade":[48],"significant":[49],"efficiency":[50],"marginal":[52],"gains.To":[54],"improve":[55],"LLM":[56],"while":[58,144],"maintaining":[59,145],"efficiency,":[60],"propose":[62],"Speculative":[64],"Reward":[65],"Model":[66],"(SRM),":[67],"a":[68,94,152,155],"plug-and-play":[69],"framework":[70,141],"seamlessly":[72],"integrates":[73],"with":[74],"strategies.Specifically,":[77],"SRM":[78,113,132],"employs":[79],"an":[80],"external":[81],"reward":[82],"assigner":[83],"predict":[85],"optimal":[86],"actions,":[87],"reducing":[88],"reliance":[89],"on":[90,114,142,184,189],"LLMs'":[91],"internal":[92],"self-evaluation.And":[93],"speculative":[95],"verification":[96],"mechanism":[97],"used":[99],"prune":[101],"suboptimal":[102],"choices":[103],"guide":[105],"toward":[108],"more":[109],"promising":[110],"steps.We":[111],"evaluate":[112],"several":[115],"complex":[116],"tasks":[118],"including":[119],"mathematical":[120],"reasoning,":[121],"planning":[122],"numerical":[124],"reasoning":[125],"specialized":[127],"domains.Experimental":[128],"results":[129],"show":[130],"reduces":[133],"costs":[134],"1/10":[136],"original":[139],"average":[143],"effectiveness.Q:":[146],"Josh":[147,182,186],"decides":[148],"try":[150],"flipping":[151],"house.He":[153],"buys":[154],"house":[156,171,213,222],"$80,000":[158],"then":[160],"puts":[161],"$50,000":[163,188],"repairs.This":[165],"increased":[166,223],"value":[168,210,219],"by":[172,224],"150%.How":[173],"much":[174,180,207],"profit":[175],"did":[176,181,208],"he":[177,196],"make?Q1:":[178],"How":[179,206],"spend":[183],"repairs?A1:":[185,217],"spent":[187],"repairs":[190],"because":[191],"was":[193],"amount":[195],"put":[197],"into":[198],"fixing":[199],"up":[200],"house.The":[202],"answer":[203,229],"$50,000.Q1:":[205],"increase":[214],"after":[215,226],"The":[218],"$120,000":[225],"repairs.The":[228],"$120,000.":[231]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}