{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T17:59:27Z","timestamp":1777658367212,"version":"3.51.4"},"reference-count":31,"publisher":"IEEE","license":[{"start":{"date-parts":[[2023,6,4]],"date-time":"2023-06-04T00:00:00Z","timestamp":1685836800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2023,6,4]],"date-time":"2023-06-04T00:00:00Z","timestamp":1685836800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2023,6,4]]},"DOI":"10.1109\/icassp49357.2023.10096570","type":"proceedings-article","created":{"date-parts":[[2023,5,5]],"date-time":"2023-05-05T17:28:30Z","timestamp":1683307710000},"page":"1-5","source":"Crossref","is-referenced-by-count":13,"title":["Training Large-Vocabulary Neural Language Models by Private Federated Learning for Resource-Constrained Devices"],"prefix":"10.1109","author":[{"given":"Mingbin","family":"Xu","sequence":"first","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Congzheng","family":"Song","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ye","family":"Tian","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Neha","family":"Agrawal","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Filip","family":"Granqvist","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Rogier","family":"van Dalen","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiao","family":"Zhang","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Arturo","family":"Argueta","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shiyi","family":"Han","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yaqiao","family":"Deng","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Leo","family":"Liu","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Anmol","family":"Walia","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Alex","family":"Jin","sequence":"additional","affiliation":[{"name":"Apple"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref13","article-title":"The fixed-size ordinally-forgetting encoding method for neural network language models","author":"hui jiang","year":"2015","journal-title":"ACL"},{"key":"ref12","author":"ramaswamy","year":"2020","journal-title":"Training production language models without memorizing user data"},{"key":"ref15","article-title":"Attention is all you need","author":"vaswani","year":"2017","journal-title":"NeurIPS"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.8.1735"},{"key":"ref31","article-title":"Large language models can be strong differentially private learners","author":"li","year":"2022","journal-title":"ICLRE"},{"key":"ref30","article-title":"Differentially private fine-tuning of language models","author":"yu","year":"2022","journal-title":"ICLRE"},{"key":"ref11","author":"ramaswamy","year":"2019","journal-title":"Federated learning for emoji prediction in a mobile keyboard"},{"key":"ref10","author":"hard","year":"2018","journal-title":"Federated learning for mobile keyboard prediction"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-1767"},{"key":"ref1","article-title":"Communication-efficient learning of deep networks from decentralized data","author":"mcmahan","year":"2017","journal-title":"AISTATS"},{"key":"ref17","article-title":"Learning differentially private recurrent language models","author":"mcmahan","year":"2018","journal-title":"ICLRE"},{"key":"ref16","article-title":"Adaptive federated optimization","author":"reddi","year":"2021","journal-title":"ICLRE"},{"key":"ref19","article-title":"Rényi differential privacy of the sampled gaussian mechanism","author":"mironov","year":"2019","journal-title":"CoRR"},{"key":"ref18","first-page":"263","article-title":"Rényi differential privacy","author":"mironov","year":"2017","journal-title":"CSF"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/E17-2025"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2016.7472805"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053973"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1006\/csla.2001.0184"},{"key":"ref20","article-title":"Lora: Low-rank adaptation of large language models","author":"hu","year":"2021","journal-title":"CoRR"},{"key":"ref22","article-title":"A fast and simple algorithm for training neural probabilistic language models","author":"mnih","year":"2012","journal-title":"ICML"},{"key":"ref21","article-title":"Noise-contrastive estimation: A new estimation principle for unnormalized statistical models","author":"gutmann","year":"2010","journal-title":"AISTATS"},{"key":"ref28","article-title":"Bert: Pre-training of deep bidirectional transformers for language understanding","author":"devlin","year":"2019","journal-title":"NAACL"},{"key":"ref27","article-title":"Measuring the effects of non-identical data distribution for federated visual classification","author":"hsu","year":"2019","journal-title":"CoRR"},{"key":"ref29","author":"anil","year":"2021","journal-title":"Large-scale differentially private bert"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1145\/2976749.2978318"},{"key":"ref7","article-title":"Calibrating noise to sensitivity in private data analysis","author":"dwork","year":"2006","journal-title":"TCC"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/FOCS.2014.56"},{"key":"ref4","article-title":"Deep leakage from gradients","author":"zhu","year":"2019","journal-title":"NeurIPS"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/SP.2019.00029"},{"key":"ref6","article-title":"Extracting training data from large language models","author":"carlini","year":"2021","journal-title":"Usenix Security"},{"key":"ref5","article-title":"The secret sharer: Evaluating and testing unintended memorization in neural networks","author":"carlini","year":"2019","journal-title":"Usenix Security"}],"event":{"name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","location":"Rhodes Island, Greece","start":{"date-parts":[[2023,6,4]]},"end":{"date-parts":[[2023,6,10]]}},"container-title":["ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/10094559\/10094560\/10096570.pdf?arnumber=10096570","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,11,20]],"date-time":"2023-11-20T18:57:55Z","timestamp":1700506675000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10096570\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,6,4]]},"references-count":31,"URL":"https:\/\/doi.org\/10.1109\/icassp49357.2023.10096570","relation":{},"subject":[],"published":{"date-parts":[[2023,6,4]]}}}