{"id":"https://openalex.org/W7140085809","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.117","title":"PIRA: Preference-Oriented Instruction-Tuned Reward Models with Dual Aggregation","display_name":"PIRA: Preference-Oriented Instruction-Tuned Reward Models with Dual Aggregation","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7140085809","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.117"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.findings-eacl.117","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.117","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.117","pdf_url":"https://aclanthology.org/2026.findings-eacl.117.pdf","source":{"id":"https://openalex.org/S7407087051","display_name":"Findings of the Association for Computational Linguistics: EACL","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"Findings of the Association for Computational Linguistics: EACL 2026","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2026.findings-eacl.117.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134375671","display_name":"Yongfu Xue","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Yongfu Xue","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5134375671"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.14148656,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"2226","last_page":"2234"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.4514000117778778,"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":0.4514000117778778,"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":0.2085999995470047,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.05689999833703041,"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/reward-models","display_name":"reward models","score":0.7400000095367432},{"id":"https://openalex.org/keywords/large-language-model-alignment","display_name":"large language model alignment","score":0.7789999842643738}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4643000066280365},{"id":"https://openalex.org/C2780980858","wikidata":"https://www.wikidata.org/wiki/Q110022","display_name":"Dual (grammatical number)","level":2,"score":0.385699987411499},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.37770000100135803},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32510000467300415},{"id":"https://openalex.org/C189950617","wikidata":"https://www.wikidata.org/wiki/Q937228","display_name":"Property (philosophy)","level":2,"score":0.3174000084400177},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.30379998683929443},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.2680000066757202},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.25609999895095825},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.24860000610351562},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.23669999837875366}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2026.findings-eacl.117","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.117","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.117","pdf_url":"https://aclanthology.org/2026.findings-eacl.117.pdf","source":{"id":"https://openalex.org/S7407087051","display_name":"Findings of the Association for Computational Linguistics: EACL","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"Findings of the Association for Computational Linguistics: EACL 2026","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2511.20668","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2511.20668","pdf_url":"https://arxiv.org/pdf/2511.20668","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.2511.20668","doi":"https://doi.org/10.48550/arxiv.2511.20668","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2511.20668","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":"doi:10.18653/v1/2026.findings-eacl.117","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.117","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.117","pdf_url":"https://aclanthology.org/2026.findings-eacl.117.pdf","source":{"id":"https://openalex.org/S7407087051","display_name":"Findings of the Association for Computational Linguistics: EACL","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"Findings of the Association for Computational Linguistics: EACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7140085809.pdf","grobid_xml":"https://content.openalex.org/works/W7140085809.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7166827766","https://openalex.org/W4403365300","https://openalex.org/W7138363763","https://openalex.org/W7214106624","https://openalex.org/W7134031930","https://openalex.org/W4410314353","https://openalex.org/W4415335523","https://openalex.org/W4404822917","https://openalex.org/W4417299754","https://openalex.org/W6892041672"],"abstract_inverted_index":{"Reward":[0],"models":[1,20,41],"are":[2,42],"pivotal":[3],"for":[4,98],"aligning":[5],"Large":[6],"Language":[7],"Models":[8],"(LLMs)":[9],"with":[10],"human":[11],"preferences.Existing":[12],"approaches":[13],"face":[14],"two":[15],"key":[16],"limitations:":[17],"Discriminative":[18],"reward":[19,38,40,46,54,125,141],"require":[21],"large-scale":[22],"annotated":[23],"data,":[24],"as":[25],"they":[26],"cannot":[27],"exploit":[28,50],"the":[29,53,91,116],"preference":[30,86],"instruction-following":[31,87],"capability":[32],"of":[33,57],"LLMs":[34,49],"available":[35],"to":[36,45,71,82,123],"generative":[37],"models.Moreover,":[39],"particularly":[43],"prone":[44],"overoptimization,":[47],"where":[48],"weaknesses":[51],"in":[52],"function":[55],"instead":[56],"improving":[58],"true":[59],"alignment.We":[60],"introduce":[61],"PIRA,":[62],"a":[63],"training":[64],"paradigm":[65],"that":[66,131],"integrates":[67],"three":[68],"complementary":[69],"strategies":[70],"address":[72],"these":[73],"challenges:":[74],"(1)":[75],"reformulating":[76],"question-answer":[77],"pairs":[78],"into":[79],"preference-task":[80],"instructions":[81,97],"explicitly":[83],"leverage":[84],"LLMs'":[85],"capability,":[88],"(2)":[89],"averaging":[90,113],"rewards":[92],"aggregated":[93],"from":[94,115],"diverse":[95],"preferencetask":[96],"each":[99],"sample,":[100],"which":[101],"mitigates":[102,140],"task-specific":[103],"bias":[104],"and":[105,111,138],"enhances":[106,136],"robustness":[107],"across":[108],"evaluation":[109],"perspectives,":[110],"(3)":[112],"outputs":[114],"value":[117],"head":[118],"under":[119],"different":[120],"dropout":[121],"rates":[122],"stabilize":[124],"estimation.Experiments":[126],"on":[127],"public":[128],"datasets":[129],"show":[130],"PIRA":[132],"improves":[133],"performance":[134],"considerably,":[135],"generalization,":[137],"effectively":[139],"overoptimization.":[142]},"counts_by_year":[],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2026-02-06T00:00:00"}