{"id":"https://openalex.org/W4404133752","doi":"https://doi.org/10.1145/3649329.3655953","title":"Token-Picker: Accelerating Attention in Text Generation with Minimized Memory Transfer via Probability Estimation","display_name":"Token-Picker: Accelerating Attention in Text Generation with Minimized Memory Transfer via Probability Estimation","publication_year":2024,"publication_date":"2024-06-23","ids":{"openalex":"https://openalex.org/W4404133752","doi":"https://doi.org/10.1145/3649329.3655953"},"language":"en","primary_location":{"id":"doi:10.1145/3649329.3655953","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3649329.3655953","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 61st ACM/IEEE Design Automation Conference","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/A5106404793","display_name":"Junyoung Park","orcid":"https://orcid.org/0009-0005-7722-8702"},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]},{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Junyoung Park","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Republic of Korea","System LSI, Samsung Electronics, Hwaseong, Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Republic of Korea","institution_ids":["https://openalex.org/I157485424"]},{"raw_affiliation_string":"System LSI, Samsung Electronics, Hwaseong, Republic of Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047372993","display_name":"Myeonggu Kang","orcid":"https://orcid.org/0000-0003-3557-8526"},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Myeonggu Kang","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Republic of Korea","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036617506","display_name":"Yunki Han","orcid":"https://orcid.org/0000-0003-0432-9324"},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Yunki Han","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology, Daejeon, Chungcheongnam-do, Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology, Daejeon, Chungcheongnam-do, Republic of Korea","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089909294","display_name":"Yang-Gon Kim","orcid":"https://orcid.org/0009-0009-6165-6234"},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]},{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Yang-Gon Kim","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Republic of Korea","System LSI, Samsung Electronics, Hwaseong, Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Republic of Korea","institution_ids":["https://openalex.org/I157485424"]},{"raw_affiliation_string":"System LSI, Samsung Electronics, Hwaseong, Republic of Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5072316644","display_name":"Jaekang Shin","orcid":"https://orcid.org/0000-0002-5943-1599"},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]},{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jaekang Shin","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Daejeon, Republic of Korea","System LSI, Samsung Electronics, Hwaseong, Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Daejeon, Republic of Korea","institution_ids":["https://openalex.org/I157485424"]},{"raw_affiliation_string":"System LSI, Samsung Electronics, Hwaseong, Republic of Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5052390471","display_name":"Lee\u2010Sup Kim","orcid":"https://orcid.org/0000-0001-9585-4591"},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Lee-Sup Kim","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Republic of Korea"],"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology, Daejeon, Daejeon, Republic of Korea","institution_ids":["https://openalex.org/I157485424"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5106404793"],"corresponding_institution_ids":["https://openalex.org/I157485424","https://openalex.org/I2250650973"],"apc_list":null,"apc_paid":null,"fwci":2.4513,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.90835209,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"6"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9977999925613403,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9977999925613403,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10028","display_name":"Topic Modeling","score":0.9873999953269958,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9667999744415283,"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/security-token","display_name":"Security token","score":0.7616572976112366},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.74642413854599},{"id":"https://openalex.org/keywords/transfer","display_name":"Transfer (computing)","score":0.5917792320251465},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.15493783354759216},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.15230798721313477}],"concepts":[{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7616572976112366},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.74642413854599},{"id":"https://openalex.org/C2776175482","wikidata":"https://www.wikidata.org/wiki/Q1195816","display_name":"Transfer (computing)","level":2,"score":0.5917792320251465},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.15493783354759216},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.15230798721313477}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3649329.3655953","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3649329.3655953","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 61st ACM/IEEE Design Automation Conference","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1317411944","display_name":null,"funder_award_id":"IITP-2020-0-01847","funder_id":"https://openalex.org/F4320335489","funder_display_name":"Institute for Information and Communications Technology Promotion"},{"id":"https://openalex.org/G2177770941","display_name":null,"funder_award_id":"4120200113769","funder_id":"https://openalex.org/F4320322120","funder_display_name":"National Research Foundation of Korea"}],"funders":[{"id":"https://openalex.org/F4320322120","display_name":"National Research Foundation of Korea","ror":"https://ror.org/013aysd81"},{"id":"https://openalex.org/F4320335489","display_name":"Institute for Information and Communications Technology Promotion","ror":"https://ror.org/01g0hqq23"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":4,"referenced_works":["https://openalex.org/W2725159389","https://openalex.org/W3006586535","https://openalex.org/W4210401038","https://openalex.org/W4245873525"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W4388335561","https://openalex.org/W2970530566","https://openalex.org/W4288261899","https://openalex.org/W4307309205","https://openalex.org/W2967478618","https://openalex.org/W4385009901","https://openalex.org/W4385572700"],"abstract_inverted_index":{"The":[0],"attention":[1,41],"mechanism":[2],"in":[3,35,43],"text":[4],"generation":[5],"is":[6],"memory-bounded":[7],"due":[8],"to":[9,87],"its":[10],"sequential":[11],"characteristics.":[12],"Therefore,":[13],"off-chip":[14,77],"memory":[15,84],"accesses":[16],"should":[17],"be":[18],"minimized":[19],"for":[20],"faster":[21],"execution.":[22],"Although":[23],"previous":[24],"methods":[25],"addressed":[26],"this":[27],"by":[28],"pruning":[29,64],"unimportant":[30],"tokens,":[31],"they":[32],"fall":[33],"short":[34],"selectively":[36],"removing":[37,56],"tokens":[38,59],"with":[39],"near-zero":[40],"probabilities":[42],"each":[44],"instance.":[45],"Our":[46,79],"method":[47],"estimates":[48],"the":[49,52],"probability":[50,58],"before":[51],"softmax":[53],"function,":[54],"effectively":[55],"low":[57],"and":[60,92],"achieving":[61],"an":[62,88],"12.1x":[63],"ratio":[65],"without":[66],"fine-tuning.":[67],"Additionally,":[68],"we":[69],"present":[70],"a":[71,93],"hardware":[72],"design":[73],"supporting":[74],"seamless":[75],"on-demand":[76],"access.":[78],"approach":[80],"shows":[81],"2.6x":[82],"reduced":[83],"accesses,":[85],"leading":[86],"average":[89],"2.3x":[90],"speedup":[91],"2.4x":[94],"energy":[95],"efficiency.":[96]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":3}],"updated_date":"2026-03-28T08:17:26.163206","created_date":"2025-10-10T00:00:00"}
