{"id":"https://openalex.org/W1917432393","doi":"https://doi.org/10.1109/icassp.2015.7179001","title":"Scaling recurrent neural network language models","display_name":"Scaling recurrent neural network language models","publication_year":2015,"publication_date":"2015-04-01","ids":{"openalex":"https://openalex.org/W1917432393","doi":"https://doi.org/10.1109/icassp.2015.7179001","mag":"1917432393"},"language":"en","primary_location":{"id":"doi:10.1109/icassp.2015.7179001","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp.2015.7179001","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","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/A5075617282","display_name":"Will Williams","orcid":null},"institutions":[{"id":"https://openalex.org/I4210131327","display_name":"Speechmatics (United Kingdom)","ror":"https://ror.org/03dtb8870","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210131327"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Will Williams","raw_affiliation_strings":["Cantab Research, Cambridge, UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cantab Research, Cambridge, UK","institution_ids":["https://openalex.org/I4210131327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5009328512","display_name":"Niranjani Prasad","orcid":"https://orcid.org/0000-0002-6677-280X"},"institutions":[{"id":"https://openalex.org/I4210131327","display_name":"Speechmatics (United Kingdom)","ror":"https://ror.org/03dtb8870","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210131327"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Niranjani Prasad","raw_affiliation_strings":["Cantab Research, Cambridge, UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cantab Research, Cambridge, UK","institution_ids":["https://openalex.org/I4210131327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023148684","display_name":"David Mrva","orcid":null},"institutions":[{"id":"https://openalex.org/I4210131327","display_name":"Speechmatics (United Kingdom)","ror":"https://ror.org/03dtb8870","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210131327"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"David Mrva","raw_affiliation_strings":["Cantab Research, Cambridge, UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cantab Research, Cambridge, UK","institution_ids":["https://openalex.org/I4210131327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036305150","display_name":"Tom Ash","orcid":null},"institutions":[{"id":"https://openalex.org/I4210131327","display_name":"Speechmatics (United Kingdom)","ror":"https://ror.org/03dtb8870","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210131327"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Tom Ash","raw_affiliation_strings":["Cantab Research, Cambridge, UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cantab Research, Cambridge, UK","institution_ids":["https://openalex.org/I4210131327"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5103532312","display_name":"Tony Robinson","orcid":null},"institutions":[{"id":"https://openalex.org/I4210131327","display_name":"Speechmatics (United Kingdom)","ror":"https://ror.org/03dtb8870","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210131327"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Tony Robinson","raw_affiliation_strings":["Cantab Research, Cambridge, UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cantab Research, Cambridge, UK","institution_ids":["https://openalex.org/I4210131327"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I4210131327"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":49,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"5391","last_page":"5395"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9998000264167786,"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":0.9998000264167786,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9998000264167786,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9979000091552734,"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/language-model","display_name":"Language model","score":0.8533509373664856},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.803457498550415},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.6959928274154663},{"id":"https://openalex.org/keywords/recurrent-neural-network","display_name":"Recurrent neural network","score":0.6954945921897888},{"id":"https://openalex.org/keywords/machine-translation","display_name":"Machine translation","score":0.6914800405502319},{"id":"https://openalex.org/keywords/word","display_name":"Word (group theory)","score":0.6648042798042297},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.6476784944534302},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5934088826179504},{"id":"https://openalex.org/keywords/word-error-rate","display_name":"Word error rate","score":0.5830336809158325},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5490775108337402},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.5105994343757629},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.47840580344200134},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4732680320739746},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.32218682765960693},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.07874289155006409},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.06801813840866089},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.06268748641014099}],"concepts":[{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.8533509373664856},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.803457498550415},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.6959928274154663},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.6954945921897888},{"id":"https://openalex.org/C203005215","wikidata":"https://www.wikidata.org/wiki/Q79798","display_name":"Machine translation","level":2,"score":0.6914800405502319},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.6648042798042297},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.6476784944534302},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5934088826179504},{"id":"https://openalex.org/C40969351","wikidata":"https://www.wikidata.org/wiki/Q3516228","display_name":"Word error rate","level":2,"score":0.5830336809158325},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5490775108337402},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5105994343757629},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.47840580344200134},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4732680320739746},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.32218682765960693},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.07874289155006409},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.06801813840866089},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.06268748641014099},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.0},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp.2015.7179001","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp.2015.7179001","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.47999998927116394,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":28,"referenced_works":["https://openalex.org/W179875071","https://openalex.org/W196214544","https://openalex.org/W1926502259","https://openalex.org/W2008398646","https://openalex.org/W2097732278","https://openalex.org/W2108563286","https://openalex.org/W2110871230","https://openalex.org/W2117278770","https://openalex.org/W2120480077","https://openalex.org/W2131774270","https://openalex.org/W2134800885","https://openalex.org/W2145431022","https://openalex.org/W2250357346","https://openalex.org/W2250653840","https://openalex.org/W2251502575","https://openalex.org/W2271177914","https://openalex.org/W2474824677","https://openalex.org/W2611669587","https://openalex.org/W2951714314","https://openalex.org/W6607333740","https://openalex.org/W6607974698","https://openalex.org/W6632248436","https://openalex.org/W6674387193","https://openalex.org/W6677328538","https://openalex.org/W6679855610","https://openalex.org/W6691509046","https://openalex.org/W6691892052","https://openalex.org/W6720877245"],"related_works":["https://openalex.org/W2160451571","https://openalex.org/W2463396630","https://openalex.org/W2794347674","https://openalex.org/W1566315437","https://openalex.org/W2151348424","https://openalex.org/W4221142855","https://openalex.org/W2594897229","https://openalex.org/W4290708361","https://openalex.org/W2050138804","https://openalex.org/W767271433"],"abstract_inverted_index":{"This":[0],"paper":[1],"investigates":[2],"the":[3,25,65,85,90],"scaling":[4],"properties":[5],"of":[6,27,76],"Recurrent":[7],"Neural":[8],"Network":[9],"Language":[10],"Models":[11],"(RNNLMs).":[12],"We":[13,63,82],"discuss":[14],"how":[15,28],"to":[16,33,50],"train":[17,64],"very":[18],"large":[19],"RNNs":[20,68],"on":[21,57,78,89,102],"GPUs":[22],"and":[23,40,69,105],"address":[24],"questions":[26],"RNNLMs":[29,52],"scale":[30],"with":[31],"respect":[32],"model":[34],"size,":[35,37],"training-set":[36],"computational":[38],"costs":[39],"memory.":[41],"Our":[42],"analysis":[43],"shows":[44],"that":[45],"despite":[46],"being":[47],"more":[48],"costly":[49],"train,":[51],"obtain":[53],"much":[54],"lower":[55],"perplexities":[56,88],"standard":[58],"benchmarks":[59],"than":[60],"n-gram":[61],"models.":[62],"largest":[66],"known":[67],"present":[70,84],"relative":[71,108],"word":[72,94,113],"error":[73],"rates":[74],"gains":[75],"18%":[77],"an":[79],"ASR":[80],"task.":[81],"also":[83],"new":[86],"lowest":[87],"recently":[91],"released":[92],"billion":[93],"language":[95],"modelling":[96],"benchmark,":[97],"1":[98],"BLEU":[99],"point":[100],"gain":[101,111],"machine":[103],"translation":[104],"a":[106],"17%":[107],"hit":[109],"rate":[110],"in":[112],"prediction.":[114]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2023,"cited_by_count":2},{"year":2022,"cited_by_count":2},{"year":2021,"cited_by_count":3},{"year":2020,"cited_by_count":7},{"year":2019,"cited_by_count":4},{"year":2018,"cited_by_count":6},{"year":2017,"cited_by_count":7},{"year":2016,"cited_by_count":13},{"year":2015,"cited_by_count":3}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}