{"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/tying-word-vectors-and-word-classifiers-a","title":"Tying Word Vectors and Word Classifiers: A Loss Framework for Language Modeling","arxiv_id":"1611.01462","date":"2016-11-04","proceeding":null,"authors":["Hakan Inan","Khashayar Khosravi","Richard Socher"],"abstract":"Recurrent neural networks have been very successful at predicting sequences\nof words in tasks such as language modeling. However, all such models are based\non the conventional classification framework, where the model is trained\nagainst one-hot targets, and each word is represented both as an input and as\nan output in isolation. This causes inefficiencies in learning both in terms of\nutilizing all of the information and in terms of the number of parameters\nneeded to train. We introduce a novel theoretical framework that facilitates\nbetter learning in language modeling, and show that our framework leads to\ntying together the input embedding and the output projection matrices, greatly\nreducing the number of trainable variables. Our framework leads to state of the\nart performance on the Penn Treebank with a variety of network models.","url_abs":"http://arxiv.org/abs/1611.01462v3","url_pdf":"http://arxiv.org/pdf/1611.01462v3.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":"tying-word-vectors-and-word-classifiers-a","repo_url":"https://github.com/InnerPeace-Wu/im2p-tensorflow","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"tying-word-vectors-and-word-classifiers-a","repo_url":"https://github.com/JianGoForIt/YellowFin_Pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"tying-word-vectors-and-word-classifiers-a","repo_url":"https://github.com/Ravoxsg/Word-level-language-modeling","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"tying-word-vectors-and-word-classifiers-a","repo_url":"https://github.com/floydhub/word-language-model","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"tying-word-vectors-and-word-classifiers-a","repo_url":"https://github.com/rdspring1/PyTorch_GBW_LM","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"}],"methods":[{"method_slug":"weight-tying","method_name":"Weight Tying"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/language-modelling-on-penn-treebank-word","task":"Language Modelling","dataset":"Penn Treebank (Word Level)","model":"Inan et al. (2016) - Variational RHN","rank_in_archive_order":34,"of":43,"metrics":{"Test perplexity":"66.0","Validation perplexity":"68.1"},"uses_additional_data":false},{"leaderboard":"/sota/language-modelling-on-wikitext-2","task":"Language Modelling","dataset":"WikiText-2","model":"Inan et al. (2016) - Variational LSTM (tied) (h=650) + augmented loss","rank_in_archive_order":35,"of":38,"metrics":{"Test perplexity":"87.0","Validation perplexity":"91.5"},"uses_additional_data":false},{"leaderboard":"/sota/language-modelling-on-wikitext-2","task":"Language Modelling","dataset":"WikiText-2","model":"Inan et al. (2016) - Variational LSTM (tied) (h=650)","rank_in_archive_order":36,"of":38,"metrics":{"Test perplexity":"87.7","Validation perplexity":"92.3"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1611.01462","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}