{"id":"https://openalex.org/W4404781872","doi":"https://doi.org/10.18653/v1/2024.findings-emnlp.422","title":"Intermediate Layer Distillation with the Reused Teacher Classifier: A Study on the Importance of the Classifier of Attention-based Models","display_name":"Intermediate Layer Distillation with the Reused Teacher Classifier: A Study on the Importance of the Classifier of Attention-based Models","publication_year":2024,"publication_date":"2024-01-01","ids":{"openalex":"https://openalex.org/W4404781872","doi":"https://doi.org/10.18653/v1/2024.findings-emnlp.422"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2024.findings-emnlp.422","is_oa":true,"landing_page_url":"http://dx.doi.org/10.18653/v1/2024.findings-emnlp.422","pdf_url":"https://aclanthology.org/2024.findings-emnlp.422.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":"Findings of the Association for Computational Linguistics: EMNLP 2024","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2024.findings-emnlp.422.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5062819361","display_name":"Zhang Hang","orcid":"https://orcid.org/0000-0001-9236-372X"},"institutions":[{"id":"https://openalex.org/I5023651","display_name":"McGill University","ror":"https://ror.org/01pxwe438","country_code":"CA","type":"education","lineage":["https://openalex.org/I5023651"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Hang Zhang","raw_affiliation_strings":["Department of Electrical and Computer Engineering , McGill University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering , McGill University","institution_ids":["https://openalex.org/I5023651"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056921752","display_name":"Seyyed Hasan Mozafari","orcid":"https://orcid.org/0000-0002-0360-1747"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Seyyed Hasan Mozafari","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104000437","display_name":"James J. Clark","orcid":null},"institutions":[{"id":"https://openalex.org/I5023651","display_name":"McGill University","ror":"https://ror.org/01pxwe438","country_code":"CA","type":"education","lineage":["https://openalex.org/I5023651"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"James J. Clark","raw_affiliation_strings":["Department of Electrical and Computer Engineering , McGill University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering , McGill University","institution_ids":["https://openalex.org/I5023651"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034701871","display_name":"Brett H. Meyer","orcid":"https://orcid.org/0000-0002-6650-3298"},"institutions":[{"id":"https://openalex.org/I5023651","display_name":"McGill University","ror":"https://ror.org/01pxwe438","country_code":"CA","type":"education","lineage":["https://openalex.org/I5023651"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Brett H. Meyer","raw_affiliation_strings":["Department of Electrical and Computer Engineering , McGill University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering , McGill University","institution_ids":["https://openalex.org/I5023651"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5091600002","display_name":"Warren J. Gross","orcid":"https://orcid.org/0000-0002-6226-6037"},"institutions":[{"id":"https://openalex.org/I5023651","display_name":"McGill University","ror":"https://ror.org/01pxwe438","country_code":"CA","type":"education","lineage":["https://openalex.org/I5023651"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Warren J. Gross","raw_affiliation_strings":["Department of Electrical and Computer Engineering , McGill University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering , McGill University","institution_ids":["https://openalex.org/I5023651"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"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":0.18252505,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"7203","last_page":"7212"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.32019999623298645,"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/T10320","display_name":"Neural Networks and Applications","score":0.32019999623298645,"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/T10820","display_name":"Fuzzy Logic and Control Systems","score":0.27730000019073486,"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/T11975","display_name":"Evolutionary Algorithms and Applications","score":0.2621999979019165,"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/classifier","display_name":"Classifier (UML)","score":0.774170458316803},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.715516209602356},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.608741044998169},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5460550785064697},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.5228474736213684},{"id":"https://openalex.org/keywords/chemistry","display_name":"Chemistry","score":0.11758232116699219},{"id":"https://openalex.org/keywords/chromatography","display_name":"Chromatography","score":0.11547502875328064}],"concepts":[{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.774170458316803},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.715516209602356},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.608741044998169},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5460550785064697},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5228474736213684},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.11758232116699219},{"id":"https://openalex.org/C43617362","wikidata":"https://www.wikidata.org/wiki/Q170050","display_name":"Chromatography","level":1,"score":0.11547502875328064}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2024.findings-emnlp.422","is_oa":true,"landing_page_url":"http://dx.doi.org/10.18653/v1/2024.findings-emnlp.422","pdf_url":"https://aclanthology.org/2024.findings-emnlp.422.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":"Findings of the Association for Computational Linguistics: EMNLP 2024","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2024.findings-emnlp.422","is_oa":true,"landing_page_url":"http://dx.doi.org/10.18653/v1/2024.findings-emnlp.422","pdf_url":"https://aclanthology.org/2024.findings-emnlp.422.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":"Findings of the Association for Computational Linguistics: EMNLP 2024","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.47999998927116394}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4404781872.pdf","grobid_xml":"https://content.openalex.org/works/W4404781872.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2961085424","https://openalex.org/W4306674287","https://openalex.org/W3046775127","https://openalex.org/W4394896187","https://openalex.org/W3170094116","https://openalex.org/W4386462264","https://openalex.org/W3107602296","https://openalex.org/W4364306694","https://openalex.org/W4312192474","https://openalex.org/W4283697347"],"abstract_inverted_index":{"Intermediate":[0],"Layer":[1],"Distillation":[2],"(ILD)":[3],"effectively":[4],"compresses":[5],"large-scale":[6],"pre-trained":[7,42,111,127],"language":[8],"models":[9],"(PLMs).Existing":[10],"ILD":[11,60],"methods":[12],"underestimate":[13],"the":[14,18,41,66,79,82,86,102,119,124,135],"importance":[15],"of":[16,39,65,85,126],"utilizing":[17],"teacher's":[19],"discriminative":[20],"classifier":[21,44],"and":[22],"face":[23],"challenges":[24],"in":[25],"establishing":[26],"proper":[27],"layer":[28,52],"mappings.Therefore,":[29],"we":[30],"propose":[31],"ILD-RTC,":[32],"to":[33,77,93,133],"show":[34],"that":[35],"a":[36],"straightforward":[37],"implementation":[38],"reusing":[40],"teacher":[43,68,87,128],"improves":[45],"student":[46,80,136],"performance":[47,64],"even":[48],"with":[49,95],"simple":[50],"uniform":[51],"mapping.Through":[53],"extensive":[54],"experiments,":[55],"our":[56,90,99],"method":[57,122],"outperforms":[58],"other":[59],"techniques,":[61],"maintaining":[62],"97.7%":[63],"original":[67],"BERT":[69],"base":[70],"without":[71],"additional":[72],"trainable":[73],"parameters.Projectors":[74],"are":[75],"developed":[76],"help":[78],"match":[81],"hidden":[83],"size":[84],"model,":[88],"making":[89],"ILD-RTC":[91],"applicable":[92],"students":[94,108],"different":[96],"sizes.In":[97],"addition,":[98],"technique":[100],"achieves":[101],"same":[103],"average":[104],"GLUE":[105],"score":[106],"as":[107,130],"initialized":[109],"by":[110],"LMs,":[112],"saving":[113],"over":[114],"80":[115],"cost":[116],"resulting":[117],"from":[118],"pretraining":[120],"step.Our":[121],"emphasizes":[123],"reuse":[125],"classifiers":[129],"an":[131],"alternative":[132],"pre-training":[134],"for":[137],"initialization.":[138]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}