{"id":"https://openalex.org/W2994840239","doi":null,"title":"LEARNED STEP SIZE QUANTIZATION","display_name":"LEARNED STEP SIZE QUANTIZATION","publication_year":2020,"publication_date":"2020-04-30","ids":{"openalex":"https://openalex.org/W2994840239","mag":"2994840239"},"language":"en","primary_location":{"id":"mag:2994840239","doi":null,"is_oa":false,"landing_page_url":"https://www.openreview.net/pdf?id=rkgO66VKDS","pdf_url":null,"source":{"id":"https://openalex.org/S4306419637","display_name":"International Conference on Learning Representations","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":"International Conference on Learning Representations","raw_type":null},"volume":null,"type":"conference-paper","is_traditional":true,"indexed_in":[],"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/A5055044754","display_name":"Steven K. Esser","orcid":"https://orcid.org/0009-0008-4089-0577","observed_orcids":["https://orcid.org/0009-0008-4089-0577"]},"institutions":[{"id":"https://openalex.org/I1341412227","display_name":"IBM (United States)","ror":"https://ror.org/05hh8d621","country_code":"US","type":"company","lineage":["https://openalex.org/I1341412227"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Steven K. Esser","raw_affiliation_strings":["IBM (United States), Armonk, United States"],"raw_orcid":"https://orcid.org/0009-0008-4089-0577","affiliations":[{"raw_affiliation_string":"IBM (United States), Armonk, United States","institution_ids":["https://openalex.org/I1341412227"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5059206462","display_name":"Jeffrey L. McKinstry","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1341412227","display_name":"IBM (United States)","ror":"https://ror.org/05hh8d621","country_code":"US","type":"company","lineage":["https://openalex.org/I1341412227"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jeffrey L. McKinstry","raw_affiliation_strings":["IBM (United States), Armonk, United States"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM (United States), Armonk, United States","institution_ids":["https://openalex.org/I1341412227"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045970662","display_name":"Deepika Bablani","orcid":"https://orcid.org/0000-0002-8987-8201","observed_orcids":["https://orcid.org/0000-0002-8987-8201"]},"institutions":[{"id":"https://openalex.org/I1341412227","display_name":"IBM (United States)","ror":"https://ror.org/05hh8d621","country_code":"US","type":"company","lineage":["https://openalex.org/I1341412227"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Deepika Bablani","raw_affiliation_strings":["IBM (United States), Armonk, United States"],"raw_orcid":"https://orcid.org/0000-0002-8987-8201","affiliations":[{"raw_affiliation_string":"IBM (United States), Armonk, United States","institution_ids":["https://openalex.org/I1341412227"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113665253","display_name":"Rathinakumar Appuswamy","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1341412227","display_name":"IBM (United States)","ror":"https://ror.org/05hh8d621","country_code":"US","type":"company","lineage":["https://openalex.org/I1341412227"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Rathinakumar Appuswamy","raw_affiliation_strings":["IBM (United States), Armonk, United States"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM (United States), Armonk, United States","institution_ids":["https://openalex.org/I1341412227"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5063718511","display_name":"Dharmendra S. Modha","orcid":"https://orcid.org/0009-0005-6714-8127","observed_orcids":["https://orcid.org/0009-0005-6714-8127"]},"institutions":[{"id":"https://openalex.org/I1341412227","display_name":"IBM (United States)","ror":"https://ror.org/05hh8d621","country_code":"US","type":"company","lineage":["https://openalex.org/I1341412227"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dharmendra S. Modha","raw_affiliation_strings":["IBM (United States), Armonk, United States"],"raw_orcid":"https://orcid.org/0009-0005-6714-8127","affiliations":[{"raw_affiliation_string":"IBM (United States), Armonk, United States","institution_ids":["https://openalex.org/I1341412227"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I1341412227"],"apc_list":null,"apc_paid":null,"fwci":5.9927,"has_fulltext":false,"cited_by_count":84,"citation_normalized_percentile":{"value":0.96756124,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":88,"max":100},"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9951000213623047,"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"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9951000213623047,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.0010000000474974513,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.00039999998989515007,"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/neural-network-quantization","display_name":"neural network quantization","score":0.7509999871253967},{"id":"https://openalex.org/keywords/low-precision-training","display_name":"low-precision training","score":0.8100000023841858},{"id":"https://openalex.org/keywords/gradient-estimation","display_name":"gradient estimation","score":0.42500001192092896},{"id":"https://openalex.org/keywords/gradient-scaling","display_name":"gradient scaling","score":0.8370000123977661},{"id":"https://openalex.org/keywords/energy-efficient-deep-learning","display_name":"energy-efficient deep learning","score":0.6060000061988831}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8012251853942871},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.7959872484207153},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6861162185668945},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4709499776363373},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4518851041793823},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4055030047893524},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3995063900947571},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3584405779838562},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"mag:2994840239","doi":null,"is_oa":false,"landing_page_url":"https://www.openreview.net/pdf?id=rkgO66VKDS","pdf_url":null,"source":{"id":"https://openalex.org/S4306419637","display_name":"International Conference on Learning Representations","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":"International Conference on Learning Representations","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":30,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W1821462560","https://openalex.org/W2117539524","https://openalex.org/W2160815625","https://openalex.org/W2194775991","https://openalex.org/W2195210638","https://openalex.org/W2242818861","https://openalex.org/W2264905057","https://openalex.org/W2267635276","https://openalex.org/W2276486856","https://openalex.org/W2300242332","https://openalex.org/W2314470091","https://openalex.org/W2469490737","https://openalex.org/W2518108298","https://openalex.org/W2559767995","https://openalex.org/W2560017826","https://openalex.org/W2586654419","https://openalex.org/W2606722458","https://openalex.org/W2757910899","https://openalex.org/W2769312802","https://openalex.org/W2786771851","https://openalex.org/W2787752464","https://openalex.org/W2795077673","https://openalex.org/W2884150179","https://openalex.org/W2885688430","https://openalex.org/W2886847195","https://openalex.org/W2890248397","https://openalex.org/W2893251308","https://openalex.org/W2949117887","https://openalex.org/W2963114950"],"related_works":["https://openalex.org/W4405121697","https://openalex.org/W2885688430","https://openalex.org/W2884150179","https://openalex.org/W2954582488","https://openalex.org/W4221153540","https://openalex.org/W3173567919","https://openalex.org/W2890616119","https://openalex.org/W2890248397","https://openalex.org/W2798504536","https://openalex.org/W4312600348"],"abstract_inverted_index":{"Deep":[0],"networks":[1,99],"run":[2],"with":[3,64,139],"low":[4],"precision":[5,17,30,85,150],"operations":[6],"at":[7,122],"inference":[8],"time":[9],"offer":[10],"power":[11],"and":[12,66,76,116,125,157],"space":[13],"advantages":[14],"over":[15],"high":[16,27],"alternatives,":[18],"but":[19],"need":[20],"to":[21,50,69,114],"overcome":[22],"the":[23,47,53,103,118],"challenge":[24],"of":[25,62,74,149,163],"maintaining":[26],"accuracy":[28,49],"as":[29,151],"decreases.":[31],"Here,":[32],"we":[33,109],"present":[34],"a":[35,60,111,154,160],"method":[36],"for":[37,94,153],"training":[38,165],"such":[39,131],"networks,":[40],"Learned":[41],"Step":[42],"Size":[43],"Quantization,":[44],"that":[45,77,82,132],"achieves":[46],"highest":[48],"date":[51],"on":[52],"ImageNet":[54],"dataset":[55],"when":[56],"using":[57,146],"models,":[58],"from":[59],"variety":[61],"architectures,":[63],"weights":[65,96],"activations":[67],"quantized":[68,98],"2-,":[70],"3-":[71],"or":[72],"4-bits":[73],"precision,":[75],"can":[78,134],"train":[79],"3-bit":[80],"models":[81],"reach":[83],"full":[84],"baseline":[86],"accuracy.":[87],"Our":[88],"approach":[89,144],"builds":[90],"upon":[91],"existing":[92,164],"methods":[93],"learning":[95],"in":[97,137],"by":[100],"improving":[101],"how":[102],"quantizer":[104,128],"itself":[105],"is":[106],"configured.":[107],"Specifically,":[108],"introduce":[110],"novel":[112],"means":[113],"estimate":[115],"scale":[117],"task":[119],"loss":[120],"gradient":[121],"each":[123],"weight":[124],"activation":[126],"layer's":[127],"step":[129],"size,":[130],"it":[133],"be":[135],"learned":[136],"conjunction":[138],"other":[140],"network":[141],"parameters.":[142],"This":[143],"works":[145],"different":[147],"levels":[148],"needed":[152],"given":[155],"system":[156],"requires":[158],"only":[159],"simple":[161],"modification":[162],"code.":[166]},"counts_by_year":[{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":9},{"year":2021,"cited_by_count":46},{"year":2020,"cited_by_count":25},{"year":2019,"cited_by_count":3}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2019-12-26T00:00:00"}