{"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/you-may-not-need-attention","title":"You May Not Need Attention","arxiv_id":"1810.13409","date":"2018-10-31","proceeding":null,"authors":["Ofir Press","Noah A. Smith"],"abstract":"In NMT, how far can we get without attention and without separate encoding\nand decoding? To answer that question, we introduce a recurrent neural\ntranslation model that does not use attention and does not have a separate\nencoder and decoder. Our eager translation model is low-latency, writing target\ntokens as soon as it reads the first source token, and uses constant memory\nduring decoding. It performs on par with the standard attention-based model of\nBahdanau et al. (2014), and better on long sentences.","url_abs":"http://arxiv.org/abs/1810.13409v1","url_pdf":"http://arxiv.org/pdf/1810.13409v1.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":"you-may-not-need-attention","repo_url":"https://github.com/ofirpress/YouMayNotNeedAttention","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"nmt","task_name":"NMT"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1810.13409","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}