{"id":"https://openalex.org/W4416771671","doi":"https://doi.org/10.1109/sped67700.2025.11252515","title":"Effects of Attention Head Pruning on Encoder-only Language Models for Multilingual Recipe Classification","display_name":"Effects of Attention Head Pruning on Encoder-only Language Models for Multilingual Recipe Classification","publication_year":2025,"publication_date":"2025-10-19","ids":{"openalex":"https://openalex.org/W4416771671","doi":"https://doi.org/10.1109/sped67700.2025.11252515"},"language":null,"primary_location":{"id":"doi:10.1109/sped67700.2025.11252515","is_oa":false,"landing_page_url":"https://doi.org/10.1109/sped67700.2025.11252515","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 International Conference on Speech Technology and Human-Computer Dialogue (SpeD)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120388176","display_name":"Alin-Gabriel Anghelus","orcid":null},"institutions":[{"id":"https://openalex.org/I158333966","display_name":"Technical University of Cluj-Napoca","ror":"https://ror.org/03r8nwp71","country_code":"RO","type":"education","lineage":["https://openalex.org/I158333966"]}],"countries":["RO"],"is_corresponding":true,"raw_author_name":"Alin-Gabriel Anghelus","raw_affiliation_strings":["Technical University of Cluj-Napoca,Department of Computer Science,Romania"],"affiliations":[{"raw_affiliation_string":"Technical University of Cluj-Napoca,Department of Computer Science,Romania","institution_ids":["https://openalex.org/I158333966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104201169","display_name":"Vlad-Andrei Negru","orcid":null},"institutions":[{"id":"https://openalex.org/I158333966","display_name":"Technical University of Cluj-Napoca","ror":"https://ror.org/03r8nwp71","country_code":"RO","type":"education","lineage":["https://openalex.org/I158333966"]}],"countries":["RO"],"is_corresponding":false,"raw_author_name":"Vlad-Andrei Negru","raw_affiliation_strings":["Technical University of Cluj-Napoca,Department of Computer Science,Romania"],"affiliations":[{"raw_affiliation_string":"Technical University of Cluj-Napoca,Department of Computer Science,Romania","institution_ids":["https://openalex.org/I158333966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070444955","display_name":"Camelia Lemnaru","orcid":"https://orcid.org/0000-0002-4901-9808"},"institutions":[{"id":"https://openalex.org/I158333966","display_name":"Technical University of Cluj-Napoca","ror":"https://ror.org/03r8nwp71","country_code":"RO","type":"education","lineage":["https://openalex.org/I158333966"]}],"countries":["RO"],"is_corresponding":false,"raw_author_name":"Camelia Lemnaru","raw_affiliation_strings":["Technical University of Cluj-Napoca,Department of Computer Science,Romania"],"affiliations":[{"raw_affiliation_string":"Technical University of Cluj-Napoca,Department of Computer Science,Romania","institution_ids":["https://openalex.org/I158333966"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5049230139","display_name":"Rodica Potolea","orcid":"https://orcid.org/0000-0002-7051-3691"},"institutions":[{"id":"https://openalex.org/I158333966","display_name":"Technical University of Cluj-Napoca","ror":"https://ror.org/03r8nwp71","country_code":"RO","type":"education","lineage":["https://openalex.org/I158333966"]}],"countries":["RO"],"is_corresponding":false,"raw_author_name":"Rodica Potolea","raw_affiliation_strings":["Technical University of Cluj-Napoca,Department of Computer Science,Romania"],"affiliations":[{"raw_affiliation_string":"Technical University of Cluj-Napoca,Department of Computer Science,Romania","institution_ids":["https://openalex.org/I158333966"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5120388176"],"corresponding_institution_ids":["https://openalex.org/I158333966"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.20611511,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"29","last_page":"33"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.22990000247955322,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.22990000247955322,"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/T11448","display_name":"Face recognition and analysis","score":0.12439999729394913,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.09849999845027924,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/language-model","display_name":"Language model","score":0.7075999975204468},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.7045000195503235},{"id":"https://openalex.org/keywords/recipe","display_name":"Recipe","score":0.586899995803833},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.5580999851226807},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.54339998960495},{"id":"https://openalex.org/keywords/encode","display_name":"ENCODE","score":0.421999990940094},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.41830000281333923}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7397000193595886},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.7075999975204468},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.7045000195503235},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5892000198364258},{"id":"https://openalex.org/C2778671685","wikidata":"https://www.wikidata.org/wiki/Q219239","display_name":"Recipe","level":2,"score":0.586899995803833},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5814999938011169},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.5580999851226807},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.54339998960495},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4228000044822693},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.421999990940094},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.41830000281333923},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.40700000524520874},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.32850000262260437},{"id":"https://openalex.org/C2780312720","wikidata":"https://www.wikidata.org/wiki/Q5689100","display_name":"Head (geology)","level":2,"score":0.32510000467300415},{"id":"https://openalex.org/C2985367798","wikidata":"https://www.wikidata.org/wiki/Q1346592","display_name":"Parallel corpora","level":3,"score":0.2653999924659729},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.26109999418258667},{"id":"https://openalex.org/C2779679103","wikidata":"https://www.wikidata.org/wiki/Q5251805","display_name":"Degradation (telecommunications)","level":2,"score":0.2572999894618988},{"id":"https://openalex.org/C2983448237","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Language understanding","level":2,"score":0.25060001015663147}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/sped67700.2025.11252515","is_oa":false,"landing_page_url":"https://doi.org/10.1109/sped67700.2025.11252515","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 International Conference on Speech Technology and Human-Computer Dialogue (SpeD)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":1,"referenced_works":["https://openalex.org/W4399728742"],"related_works":[],"abstract_inverted_index":{"Pruning,":[0],"as":[1,20,47,145],"a":[2,71,94,169],"method":[3],"of":[4,57,74,86,109,148],"reducing":[5],"model":[6,102],"size":[7],"and":[8],"improving":[9],"performance,":[10,103],"has":[11],"gained":[12],"increasing":[13],"traction":[14],"in":[15,62,157],"recent":[16],"years.":[17],"Models":[18],"such":[19,46],"BERT":[21],"have":[22],"been":[23],"shown":[24],"to":[25,100,123,161],"be":[26,127],"over-parametrized:":[27],"multiple":[28],"attention":[29,59,92],"heads":[30,150],"encode":[31],"the":[32,55,107,146],"same":[33],"patterns.":[34],"Additionally,":[35],"models":[36],"trained":[37],"on":[38,43,70,83,97,113],"general":[39],"data":[40],"might":[41],"underperform":[42],"domain-specific":[44],"tasks,":[45],"recipe":[48],"interpretation.":[49],"In":[50],"this":[51],"work,":[52],"we":[53,105],"explore":[54],"effects":[56],"score-based":[58],"head":[60,93],"pruning":[61,111,121],"multilingual":[63],"transformer":[64],"models.":[65],"We":[66,89,136,152],"conduct":[67],"our":[68,162],"experiments":[69],"dataset":[72],"comprised":[73],"six":[75],"Indo-European":[76],"languages":[77,165],"with":[78],"unequal":[79],"representation":[80],"across":[81],"languages,":[82],"three":[84],"tasks":[85,140],"varying":[87],"difficulty.":[88],"grant":[90],"each":[91],"score,":[95],"based":[96,112],"its":[98],"contribution":[99],"overall":[101],"then":[104],"evaluate":[106],"impact":[108],"successive":[110],"said":[114],"score.":[115],"Our":[116],"findings":[117],"suggest":[118],"that":[119,138],"substantial":[120],"(up":[122],"$80":[124],"\\%$)":[125],"can":[126],"performed":[128],"without":[129],"major":[130],"performance":[131,143,172],"loss":[132],"when":[133,174],"applied":[134],"post-finetuning.":[135],"show":[137,141],"easier":[139],"slower":[142],"degradation":[144,173],"percentage":[147],"pruned":[149],"increases.":[151],"also":[153],"report":[154],"consistent":[155],"reductions":[156],"inference":[158],"time.":[159],"Contrary":[160],"expectations,":[163],"low-resource":[164],"did":[166],"not":[167],"suffer":[168],"significantly":[170],"faster":[171],"pruning.":[175]},"counts_by_year":[],"updated_date":"2026-03-07T16:01:11.037858","created_date":"2025-11-27T00:00:00"}