{"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/gated-orthogonal-recurrent-units-on-learning","title":"Gated Orthogonal Recurrent Units: On Learning to Forget","arxiv_id":"1706.02761","date":"2017-06-08","proceeding":null,"authors":["Li Jing","Caglar Gulcehre","John Peurifoy","Yichen Shen","Max Tegmark","Marin Soljačić","Yoshua Bengio"],"abstract":"We present a novel recurrent neural network (RNN) based model that combines\nthe remembering ability of unitary RNNs with the ability of gated RNNs to\neffectively forget redundant/irrelevant information in its memory. We achieve\nthis by extending unitary RNNs with a gating mechanism. Our model is able to\noutperform LSTMs, GRUs and Unitary RNNs on several long-term dependency\nbenchmark tasks. We empirically both show the orthogonal/unitary RNNs lack the\nability to forget and also the ability of GORU to simultaneously remember long\nterm dependencies while forgetting irrelevant information. This plays an\nimportant role in recurrent neural networks. We provide competitive results\nalong with an analysis of our model on many natural sequential tasks including\nthe bAbI Question Answering, TIMIT speech spectrum prediction, Penn TreeBank,\nand synthetic tasks that involve long-term dependencies such as algorithmic,\nparenthesis, denoising and copying tasks.","url_abs":"http://arxiv.org/abs/1706.02761v3","url_pdf":"http://arxiv.org/pdf/1706.02761v3.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":"gated-orthogonal-recurrent-units-on-learning","repo_url":"https://github.com/jingli9111/GORU-tensorflow","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"question-answering","task_name":"Question Answering"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/question-answering-on-babi","task":"Question Answering","dataset":"bAbi","model":"GORU","rank_in_archive_order":9,"of":14,"metrics":{"Accuracy (trained on 1k)":"60%"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1706.02761","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}