{"id":"https://openalex.org/W7140106362","doi":"https://doi.org/10.18653/v1/2026.eacl-long.7","title":"Beyond Sample-Level Feedback: Using Reference-Level Feedback to Guide Data Synthesis","display_name":"Beyond Sample-Level Feedback: Using Reference-Level Feedback to Guide Data Synthesis","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7140106362","doi":"https://doi.org/10.18653/v1/2026.eacl-long.7"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.eacl-long.7","doi":"https://doi.org/10.18653/v1/2026.eacl-long.7","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.7","pdf_url":"https://aclanthology.org/2026.eacl-long.7.pdf","source":{"id":"https://openalex.org/S4306418011","display_name":"Conference of the European Chapter of the Association for Computational Linguistics","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":"Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","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.eacl-long.7.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5116211062","display_name":"Shuhaib Mehri","orcid":"https://orcid.org/0009-0005-1330-3515","observed_orcids":["https://orcid.org/0009-0005-1330-3515"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shuhaib Mehri","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130396207","display_name":"Xiusi Chen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xiusi Chen","raw_affiliation_strings":["University of Illinois Urbana-Champaign"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Illinois Urbana-Champaign","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130397083","display_name":"Heng Ji","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Heng Ji","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139739701","display_name":"Dilek Hakkani-T\u00fcr","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dilek Hakkani-T\u00fcr","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"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":0.14901388,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"141","last_page":"164"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.8320000171661377,"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":0.8320000171661377,"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":0.0949999988079071,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.02590000070631504,"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/synthetic-data-generation","display_name":"synthetic data generation","score":0.746999979019165},{"id":"https://openalex.org/keywords/data-synthesis","display_name":"data synthesis","score":0.5789999961853027},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.7739999890327454},{"id":"https://openalex.org/keywords/data-quality","display_name":"data quality","score":0.4519999921321869},{"id":"https://openalex.org/keywords/llama-3-1-8b-instruct","display_name":"Llama-3.1-8B-Instruct","score":0.8309999704360962},{"id":"https://openalex.org/keywords/mistral-7b-instruct","display_name":"Mistral-7B-Instruct","score":0.9399999976158142},{"id":"https://openalex.org/keywords/alpacaeval","display_name":"AlpacaEval","score":0.8809999823570251}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.571399986743927},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.29589998722076416},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.28850001096725464},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2702000141143799},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.26980000734329224},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2623000144958496},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.25769999623298645},{"id":"https://openalex.org/C17500928","wikidata":"https://www.wikidata.org/wiki/Q959968","display_name":"Control system","level":2,"score":0.2529999911785126},{"id":"https://openalex.org/C2778348673","wikidata":"https://www.wikidata.org/wiki/Q739302","display_name":"Production (economics)","level":2,"score":0.2475000023841858},{"id":"https://openalex.org/C133462117","wikidata":"https://www.wikidata.org/wiki/Q4929239","display_name":"Data collection","level":2,"score":0.23960000276565552}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2026.eacl-long.7","doi":"https://doi.org/10.18653/v1/2026.eacl-long.7","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.7","pdf_url":"https://aclanthology.org/2026.eacl-long.7.pdf","source":{"id":"https://openalex.org/S4306418011","display_name":"Conference of the European Chapter of the Association for Computational Linguistics","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":"Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2502.04511","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2502.04511","pdf_url":"https://arxiv.org/pdf/2502.04511","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2502.04511","doi":"https://doi.org/10.48550/arxiv.2502.04511","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2502.04511","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.eacl-long.7","doi":"https://doi.org/10.18653/v1/2026.eacl-long.7","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.7","pdf_url":"https://aclanthology.org/2026.eacl-long.7.pdf","source":{"id":"https://openalex.org/S4306418011","display_name":"Conference of the European Chapter of the Association for Computational Linguistics","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":"Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7140106362.pdf","grobid_xml":"https://content.openalex.org/works/W7140106362.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7126207509","https://openalex.org/W4400910426","https://openalex.org/W6948249650","https://openalex.org/W4403570649","https://openalex.org/W4415320100","https://openalex.org/W4387947646","https://openalex.org/W4412888066","https://openalex.org/W4412944565","https://openalex.org/W4414875567","https://openalex.org/W4388032353"],"abstract_inverted_index":{"High-quality":[0],"instruction-tuning":[1],"data":[2,22,42,127],"is":[3],"crucial":[4],"for":[5,28],"developing":[6],"Large":[7],"Language":[8],"Models":[9],"(LLMs)":[10],"that":[11,58,109],"can":[12],"effectively":[13],"navigate":[14],"real-world":[15],"tasks":[16],"and":[17,87,122,125],"follow":[18],"human":[19],"instructions.While":[20],"synthetic":[21],"generation":[23],"offers":[24],"a":[25,34,56,80,100],"scalable":[26],"approach":[27],"creating":[29],"such":[30],"datasets,":[31],"it":[32],"imposes":[33],"quality":[35],"ceiling":[36],"where":[37],"models":[38],"trained":[39],"on":[40,89,104],"the":[41,45,69],"cannot":[43],"outperform":[44],"LLM":[46],"generating":[47],"it.To":[48],"overcome":[49],"this":[50,75],"limitation,":[51],"we":[52,77],"introduce":[53],"REFERENCE-LEVEL":[54,110],"FEEDBACK,":[55],"paradigm":[57],"extracts":[59],"desirable":[60],"characteristics":[61],"from":[62],"carefully":[63],"curated":[64],"reference":[65],"samples":[66],"to":[67],"guide":[68],"synthesis":[70],"of":[71,82],"higher-quality":[72],"instruction-response":[73,84],"pairs.Using":[74],"approach,":[76],"synthesize":[78],"REFED,":[79],"dataset":[81],"10K":[83],"pairs.Fine-tuning":[85],"Llama-3.1-8B-Instruct":[86],"Mistral-7B-Instruct":[88],"REFED":[90],"demonstrate":[91,108],"state-of-the-art":[92],"performance":[93],"among":[94],"similarly":[95],"sized":[96],"models,":[97],"notably":[98],"reaching":[99],"43.96%":[101],"length-controlled":[102],"winrate":[103],"AlpacaEval":[105],"2.0.Extensive":[106],"experiments":[107],"FEEDBACK":[111],"consistently":[112],"outperforms":[113],"traditional":[114],"sample-level":[115],"feedback":[116],"methods,":[117],"generalizes":[118],"across":[119],"model":[120],"architectures,":[121],"produces":[123],"high-quality":[124],"diverse":[126],"at":[128],"low":[129],"cost.":[130]},"counts_by_year":[],"updated_date":"2026-10-12T09:40:04.860316","created_date":"2026-03-24T00:00:00"}