{"id":"https://openalex.org/W4411726092","doi":"https://doi.org/10.1109/iscas56072.2025.11043162","title":"A Reinforcement Learning-Based Retraining-Free Pruning for Encoder-Based Language Models","display_name":"A Reinforcement Learning-Based Retraining-Free Pruning for Encoder-Based Language Models","publication_year":2025,"publication_date":"2025-05-25","ids":{"openalex":"https://openalex.org/W4411726092","doi":"https://doi.org/10.1109/iscas56072.2025.11043162"},"language":"en","primary_location":{"id":"doi:10.1109/iscas56072.2025.11043162","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas56072.2025.11043162","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on Circuits and Systems (ISCAS)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5036987044","display_name":"Bin Xie","orcid":"https://orcid.org/0000-0001-6286-0573"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bobin Xie","raw_affiliation_strings":["China SUEP"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China SUEP","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Renda Han","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Renda Han","raw_affiliation_strings":["China HNU"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China HNU","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065454900","display_name":"Guangzhen Yao","orcid":"https://orcid.org/0009-0002-1323-6998"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guangzhen Yao","raw_affiliation_strings":["China NENU"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China NENU","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100601765","display_name":"Haiming Li","orcid":"https://orcid.org/0000-0001-8605-6395"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Haiming Li","raw_affiliation_strings":["China SUEP"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China SUEP","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5039135711","display_name":"Simeng Zhang","orcid":"https://orcid.org/0000-0002-2884-7786"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Simeng Zhang","raw_affiliation_strings":["China NENU"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China NENU","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100415363","display_name":"Yutong Chen","orcid":"https://orcid.org/0000-0002-3033-4138"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yutong Chen","raw_affiliation_strings":["China USTC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China USTC","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100598322","display_name":"Sandong Zhu","orcid":"https://orcid.org/0009-0005-2104-9887"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sandong Zhu","raw_affiliation_strings":["China NENU"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China NENU","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5007920389","display_name":"Long Zhang","orcid":"https://orcid.org/0000-0002-8662-1830"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Long Zhang","raw_affiliation_strings":["China NENU"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China NENU","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.10641733,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9269000291824341,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9269000291824341,"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":0.9049999713897705,"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/retraining","display_name":"Retraining","score":0.8747742176055908},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.7443765997886658},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.7288821339607239},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6933454275131226},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.6132937669754028},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4857495427131653},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4527393579483032},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.32881730794906616},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.32025301456451416}],"concepts":[{"id":"https://openalex.org/C2778712577","wikidata":"https://www.wikidata.org/wiki/Q3505966","display_name":"Retraining","level":2,"score":0.8747742176055908},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7443765997886658},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.7288821339607239},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6933454275131226},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.6132937669754028},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4857495427131653},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4527393579483032},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.32881730794906616},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.32025301456451416},{"id":"https://openalex.org/C155202549","wikidata":"https://www.wikidata.org/wiki/Q178803","display_name":"International trade","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C6557445","wikidata":"https://www.wikidata.org/wiki/Q173113","display_name":"Agronomy","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iscas56072.2025.11043162","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas56072.2025.11043162","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on Circuits and Systems (ISCAS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.5400000214576721}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":19,"referenced_works":["https://openalex.org/W2251939518","https://openalex.org/W2739351760","https://openalex.org/W2923014074","https://openalex.org/W2963323070","https://openalex.org/W2963748441","https://openalex.org/W2963846996","https://openalex.org/W2979691890","https://openalex.org/W3034288893","https://openalex.org/W3106070274","https://openalex.org/W3136363192","https://openalex.org/W3174657338","https://openalex.org/W4404783751","https://openalex.org/W4406859087","https://openalex.org/W6605323724","https://openalex.org/W6754875771","https://openalex.org/W6775706467","https://openalex.org/W6810362000","https://openalex.org/W6850764752","https://openalex.org/W7062851666"],"related_works":["https://openalex.org/W2081982437","https://openalex.org/W4394857231","https://openalex.org/W2027050655","https://openalex.org/W3028244590","https://openalex.org/W4254349500","https://openalex.org/W2014369232","https://openalex.org/W3122042562","https://openalex.org/W2050078012","https://openalex.org/W2060761133","https://openalex.org/W2360307734"],"abstract_inverted_index":{"Natural":[0],"Language":[1],"Processing":[2],"(NLP)":[3],"has":[4],"achieved":[5],"significant":[6],"success":[7],"in":[8,155,169,181],"complex":[9],"tasks":[10],"across":[11],"various":[12],"domains,":[13],"yet":[14],"it":[15],"also":[16],"brings":[17],"high":[18],"computational":[19,186],"costs":[20],"and":[21,35,40,54,110,122,143],"inference":[22],"delays.":[23],"Pruning,":[24],"as":[25],"a":[26,91,160,166],"model":[27,33,46,83,153],"optimization":[28],"technique,":[29],"can":[30],"effectively":[31,123],"reduce":[32],"complexity":[34],"enhance":[36],"its":[37,179],"generalization":[38],"capability":[39],"efficiency.":[41],"However,":[42],"current":[43],"encoder-based":[44],"language":[45],"pruning":[47,93,176],"algorithms":[48],"often":[49],"lack":[50],"robust":[51],"dynamic":[52],"adaptability":[53],"tend":[55],"to":[56,75,120,173],"focus":[57],"only":[58],"on":[59,96,140],"short-term":[60],"optimal":[61],"solutions,":[62],"without":[63,127],"fully":[64],"considering":[65],"the":[66,128,141,156],"interactions":[67],"between":[68],"different":[69],"solutions.":[70],"This":[71,117],"limits":[72],"their":[73],"ability":[74],"find":[76,124],"global":[77,125],"optima,":[78],"thereby":[79],"potentially":[80],"impacting":[81],"overall":[82],"performance.":[84],"To":[85],"address":[86],"these":[87],"challenges,":[88],"we":[89],"propose":[90],"structured":[92],"algorithm":[94,118],"based":[95],"reinforcement":[97],"learning,":[98],"named":[99],"RLM":[100,150,164],"(Reinforcement":[101],"Learning":[102],"Masking),":[103],"which":[104],"includes":[105],"QLOM":[106],"(Q-Learning":[107],"Optimization":[108],"Mask)":[109],"QRMT":[111],"(Quasi-Minimal":[112],"Residual":[113],"Mask":[114],"Tuning)":[115],"components.":[116],"aims":[119],"rapidly":[121],"optima":[126],"need":[129],"for":[130],"retraining.":[131],"We":[132],"evaluated":[133],"this":[134],"method":[135],"using":[136],"BERTBASEand":[137],"DistilBERT":[138],"models":[139],"GLUE":[142],"SQuAD":[144,157],"benchmarks.":[145],"Experimental":[146],"results":[147],"show":[148],"that":[149],"significantly":[151],"enhances":[152],"accuracy":[154],"benchmark.":[158],"Under":[159],"60%":[161],"FLOPs":[162],"constraint,":[163],"achieves":[165],"8.45%":[167],"increase":[168],"F1":[170],"score":[171],"compared":[172],"existing":[174],"retraining-free":[175],"algorithms,":[177],"demonstrating":[178],"effectiveness":[180],"improving":[182],"performance":[183],"while":[184],"managing":[185],"resources":[187],"efficiently.":[188]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}