{"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/byte-level-recursive-convolutional-auto","title":"Byte-Level Recursive Convolutional Auto-Encoder for Text","arxiv_id":"1802.01817","date":"2018-02-06","proceeding":"ICLR 2018 1","authors":["Xiang Zhang","Yann Lecun"],"abstract":"This article proposes to auto-encode text at byte-level using convolutional\nnetworks with a recursive architecture. The motivation is to explore whether it\nis possible to have scalable and homogeneous text generation at byte-level in a\nnon-sequential fashion through the simple task of auto-encoding. We show that\nnon-sequential text generation from a fixed-length representation is not only\npossible, but also achieved much better auto-encoding results than recurrent\nnetworks. The proposed model is a multi-stage deep convolutional\nencoder-decoder framework using residual connections, containing up to 160\nparameterized layers. Each encoder or decoder contains a shared group of\nmodules that consists of either pooling or upsampling layers, making the\nnetwork recursive in terms of abstraction levels in representation. Results for\n6 large-scale paragraph datasets are reported, in 3 languages including Arabic,\nChinese and English. Analyses are conducted to study several properties of the\nproposed model.","url_abs":"http://arxiv.org/abs/1802.01817v1","url_pdf":"http://arxiv.org/pdf/1802.01817v1.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":"byte-level-recursive-convolutional-auto","repo_url":"https://github.com/smalik169/recursive-convolutional-autoencoder","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"text-generation","task_name":"Text Generation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}