{"id":"https://openalex.org/W4391661692","doi":"https://doi.org/10.1109/tai.2024.3363686","title":"Explainable Attention Pruning: A Metalearning-Based Approach","display_name":"Explainable Attention Pruning: A Metalearning-Based Approach","publication_year":2024,"publication_date":"2024-02-08","ids":{"openalex":"https://openalex.org/W4391661692","doi":"https://doi.org/10.1109/tai.2024.3363686"},"language":"en","primary_location":{"id":"doi:10.1109/tai.2024.3363686","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tai.2024.3363686","pdf_url":null,"source":{"id":"https://openalex.org/S4210169448","display_name":"IEEE Transactions on Artificial Intelligence","issn_l":"2691-4581","issn":["2691-4581"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Artificial Intelligence","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://hal.science/hal-04447491/document","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5030351062","display_name":"Praboda Rajapaksha","orcid":"https://orcid.org/0000-0002-6747-6367"},"institutions":[{"id":"https://openalex.org/I16038530","display_name":"Aberystwyth University","ror":"https://ror.org/015m2p889","country_code":"GB","type":"education","lineage":["https://openalex.org/I16038530"]},{"id":"https://openalex.org/I4210145102","display_name":"Institut Polytechnique de Paris","ror":"https://ror.org/042tfbd02","country_code":"FR","type":"education","lineage":["https://openalex.org/I4210145102"]},{"id":"https://openalex.org/I4405261500","display_name":"Services r\u00e9partis, Architectures, MOd\u00e9lisation, Validation, Administration des R\u00e9seaux","ror":"https://ror.org/05d29yv45","country_code":"FR","type":"facility","lineage":["https://openalex.org/I205703379","https://openalex.org/I4210145102","https://openalex.org/I4387153010","https://openalex.org/I4405261500"]}],"countries":["FR","GB"],"is_corresponding":true,"raw_author_name":"Praboda Rajapaksha","raw_affiliation_strings":["Samovar, Telecom SudParis, Institut Polytechnique de Paris, Palaiseau, France","Aberystwyth University (Penglais, Aberystwyth, Ceredigion, SY23 3FL - United Kingdom)"],"raw_orcid":"https://orcid.org/0000-0002-6747-6367","affiliations":[{"raw_affiliation_string":"Samovar, Telecom SudParis, Institut Polytechnique de Paris, Palaiseau, France","institution_ids":["https://openalex.org/I4405261500","https://openalex.org/I4210145102"]},{"raw_affiliation_string":"Aberystwyth University (Penglais, Aberystwyth, Ceredigion, SY23 3FL - United Kingdom)","institution_ids":["https://openalex.org/I16038530"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5107205316","display_name":"No\u00ebl Crespi","orcid":"https://orcid.org/0000-0003-2962-192X"},"institutions":[{"id":"https://openalex.org/I4210145102","display_name":"Institut Polytechnique de Paris","ror":"https://ror.org/042tfbd02","country_code":"FR","type":"education","lineage":["https://openalex.org/I4210145102"]},{"id":"https://openalex.org/I4405261500","display_name":"Services r\u00e9partis, Architectures, MOd\u00e9lisation, Validation, Administration des R\u00e9seaux","ror":"https://ror.org/05d29yv45","country_code":"FR","type":"facility","lineage":["https://openalex.org/I205703379","https://openalex.org/I4210145102","https://openalex.org/I4387153010","https://openalex.org/I4405261500"]}],"countries":["FR"],"is_corresponding":false,"raw_author_name":"Noel Crespi","raw_affiliation_strings":["Samovar, Telecom SudParis, Institut Polytechnique de Paris, Palaiseau, France"],"raw_orcid":"https://orcid.org/0000-0003-2962-192X","affiliations":[{"raw_affiliation_string":"Samovar, Telecom SudParis, Institut Polytechnique de Paris, Palaiseau, France","institution_ids":["https://openalex.org/I4405261500","https://openalex.org/I4210145102"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5030351062"],"corresponding_institution_ids":["https://openalex.org/I16038530","https://openalex.org/I4210145102","https://openalex.org/I4405261500"],"apc_list":null,"apc_paid":null,"fwci":2.8601,"has_fulltext":true,"cited_by_count":9,"citation_normalized_percentile":{"value":0.91317607,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":90,"max":99},"biblio":{"volume":"5","issue":"6","first_page":"2505","last_page":"2516"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9991999864578247,"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":0.9991999864578247,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.9991999864578247,"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":0.9958999752998352,"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/pruning","display_name":"Pruning","score":0.7797706127166748},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7395188808441162},{"id":"https://openalex.org/keywords/flops","display_name":"FLOPS","score":0.6811342239379883},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.6590456366539001},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6543224453926086},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.5482831001281738},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.48911339044570923},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.4471818506717682},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.43354976177215576}],"concepts":[{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.7797706127166748},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7395188808441162},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.6811342239379883},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.6590456366539001},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6543224453926086},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5482831001281738},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.48911339044570923},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.4471818506717682},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.43354976177215576},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.0},{"id":"https://openalex.org/C111368507","wikidata":"https://www.wikidata.org/wiki/Q43518","display_name":"Oceanography","level":1,"score":0.0},{"id":"https://openalex.org/C127313418","wikidata":"https://www.wikidata.org/wiki/Q1069","display_name":"Geology","level":0,"score":0.0},{"id":"https://openalex.org/C6557445","wikidata":"https://www.wikidata.org/wiki/Q173113","display_name":"Agronomy","level":1,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/tai.2024.3363686","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tai.2024.3363686","pdf_url":null,"source":{"id":"https://openalex.org/S4210169448","display_name":"IEEE Transactions on Artificial Intelligence","issn_l":"2691-4581","issn":["2691-4581"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Artificial Intelligence","raw_type":"journal-article"},{"id":"pmh:oai:HAL:hal-04447491v1","is_oa":true,"landing_page_url":"https://hal.science/hal-04447491","pdf_url":"https://hal.science/hal-04447491/document","source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"IEEE Transactions on Artificial Intelligence, 2024, pp.1-12. ⟨10.1109/TAI.2024.3363686⟩","raw_type":"Journal articles"}],"best_oa_location":{"id":"pmh:oai:HAL:hal-04447491v1","is_oa":true,"landing_page_url":"https://hal.science/hal-04447491","pdf_url":"https://hal.science/hal-04447491/document","source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"IEEE Transactions on Artificial Intelligence, 2024, pp.1-12. ⟨10.1109/TAI.2024.3363686⟩","raw_type":"Journal articles"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4391661692.pdf","grobid_xml":"https://content.openalex.org/works/W4391661692.grobid-xml"},"referenced_works_count":36,"referenced_works":["https://openalex.org/W1965555277","https://openalex.org/W2516809705","https://openalex.org/W2891503716","https://openalex.org/W2896457183","https://openalex.org/W2915589364","https://openalex.org/W2923014074","https://openalex.org/W2946794439","https://openalex.org/W2970120757","https://openalex.org/W2972324944","https://openalex.org/W2975381464","https://openalex.org/W3008082059","https://openalex.org/W3034368386","https://openalex.org/W3034457371","https://openalex.org/W3038012435","https://openalex.org/W3102519438","https://openalex.org/W3103754749","https://openalex.org/W3127812658","https://openalex.org/W3136363192","https://openalex.org/W3138154797","https://openalex.org/W3159727696","https://openalex.org/W3166316513","https://openalex.org/W4205952419","https://openalex.org/W4206634569","https://openalex.org/W4226126941","https://openalex.org/W4287715551","https://openalex.org/W4387966251","https://openalex.org/W6755207826","https://openalex.org/W6759263581","https://openalex.org/W6762945437","https://openalex.org/W6768080748","https://openalex.org/W6774293493","https://openalex.org/W6777017071","https://openalex.org/W6779124799","https://openalex.org/W6780835498","https://openalex.org/W6788001715","https://openalex.org/W6795651637"],"related_works":["https://openalex.org/W4315697128","https://openalex.org/W3102845713","https://openalex.org/W2971502891","https://openalex.org/W2383111961","https://openalex.org/W2365952365","https://openalex.org/W3205506801","https://openalex.org/W2352448290","https://openalex.org/W2380820513","https://openalex.org/W2913146933","https://openalex.org/W2372385138"],"abstract_inverted_index":{"Pruning,":[0],"as":[1,87,89],"a":[2,62,155],"technique":[3],"to":[4,33,48,177,189],"reduce":[5],"the":[6,78,90,104,110,117,121,159,190],"complexity":[7],"and":[8,37,70,133,162],"size":[9],"of":[10,77,106,109,127,145,158],"Transformer-based":[11],"models,":[12,86,124],"has":[13],"gained":[14],"significant":[15,140,179,192],"attention":[16,73,176],"in":[17,125,143],"recent":[18],"years.":[19],"While":[20,171],"various":[21],"models":[22,161,173],"have":[23],"been":[24],"successfully":[25],"pruned,":[26],"pruning":[27,64,97,101],"BERT":[28,50,93],"poses":[29],"unique":[30],"challenges":[31],"due":[32],"their":[34],"fine-grained":[35],"structure":[36],"overparameterization.":[38],"However,":[39,136],"by":[40],"carefully":[41],"considering":[42],"these":[43],"factors,":[44],"it":[45],"is":[46,81,138],"possible":[47],"prune":[49],"without":[51],"significantly":[52],"degrading":[53],"its":[54],"pre-trained":[55],"loss.":[56,135],"In":[57],"this":[58],"paper,":[59],"we":[60,153],"propose":[61],"Meta-learning-based":[63],"approach":[65],"that":[66,116],"can":[67],"adaptively":[68],"identify":[69],"eliminate":[71],"insignificant":[72],"weights.":[74],"The":[75,95,112],"performance":[76],"proposed":[79,118,164],"model":[80,119,165,185],"compared":[82],"with":[83],"several":[84],"baseline":[85,96,123,160],"well":[88],"default":[91],"fine-tuned":[92],"model.":[94],"strategies":[98],"employed":[99],"low-level":[100],"techniques,":[102],"targeting":[103],"removal":[105],"only":[107],"20%":[108],"connections.":[111],"experimental":[113],"results":[114],"show":[115],"outperforms":[120],"other":[122,172],"terms":[126,144],"lower":[128,134],"inference":[129],"latency,":[130],"higher":[131,187],"MCC":[132],"there":[137],"no":[139],"improvement":[141],"observed":[142],"average":[146],"FLOPs":[147],"(floating-point":[148],"operations":[149],"per":[150],"second).":[151],"Furthermore,":[152],"conduct":[154],"comparative":[156],"evaluation":[157],"our":[163,184],"using":[166],"two":[167],"explainable":[168],"(XAI)":[169],"approaches.":[170],"allocate":[174],"reasonable":[175],"less":[178],"words":[180],"for":[181],"sentiment":[182],"classification,":[183],"assigns":[186],"probabilities":[188],"most":[191],"sentimental":[193],"words.":[194]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":6},{"year":2024,"cited_by_count":1}],"updated_date":"2026-05-19T08:33:51.333923","created_date":"2025-10-10T00:00:00"}