{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/learning-longer-memory-in-recurrent-neural","title":"Learning Longer Memory in Recurrent Neural Networks","arxiv_id":"1412.7753","date":"2014-12-24","proceeding":null,"authors":["Tomas Mikolov","Armand Joulin","Sumit Chopra","Michael Mathieu","Marc'Aurelio Ranzato"],"abstract":"Recurrent neural network is a powerful model that learns temporal patterns in\nsequential data. For a long time, it was believed that recurrent networks are\ndifficult to train using simple optimizers, such as stochastic gradient\ndescent, due to the so-called vanishing gradient problem. In this paper, we\nshow that learning longer term patterns in real data, such as in natural\nlanguage, is perfectly possible using gradient descent. This is achieved by\nusing a slight structural modification of the simple recurrent neural network\narchitecture. We encourage some of the hidden units to change their state\nslowly by making part of the recurrent weight matrix close to identity, thus\nforming kind of a longer term memory. We evaluate our model in language\nmodeling experiments, where we obtain similar performance to the much more\ncomplex Long Short Term Memory (LSTM) networks (Hochreiter & Schmidhuber,\n1997).","url_abs":"http://arxiv.org/abs/1412.7753v2","url_pdf":"http://arxiv.org/pdf/1412.7753v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"learning-longer-memory-in-recurrent-neural","repo_url":"https://github.com/facebook/SCRNNs","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"torch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"learning-longer-memory-in-recurrent-neural","repo_url":"https://github.com/facebookarchive/SCRNNs","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"torch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"learning-longer-memory-in-recurrent-neural","repo_url":"https://github.com/myinnos/List-of-Machine-Learning-And-Statistics-Repositories","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"torch","reach":{"status":"ok"}},{"paper_slug":"learning-longer-memory-in-recurrent-neural","repo_url":"https://github.com/myinnos/Machine-Learning-And-Statistics-Repos","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"torch","reach":{"status":"ok"}},{"paper_slug":"learning-longer-memory-in-recurrent-neural","repo_url":"https://github.com/webgeist/scrnn-tensorflow","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1412.7753","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}