{"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/a-latent-variable-recurrent-neural-network","title":"A Latent Variable Recurrent Neural Network for Discourse Relation Language Models","arxiv_id":"1603.01913","date":"2016-03-07","proceeding":null,"authors":["Yangfeng Ji","Gholamreza Haffari","Jacob Eisenstein"],"abstract":"This paper presents a novel latent variable recurrent neural network\narchitecture for jointly modeling sequences of words and (possibly latent)\ndiscourse relations between adjacent sentences. A recurrent neural network\ngenerates individual words, thus reaping the benefits of\ndiscriminatively-trained vector representations. The discourse relations are\nrepresented with a latent variable, which can be predicted or marginalized,\ndepending on the task. The resulting model can therefore employ a training\nobjective that includes not only discourse relation classification, but also\nword prediction. As a result, it outperforms state-of-the-art alternatives for\ntwo tasks: implicit discourse relation classification in the Penn Discourse\nTreebank, and dialog act classification in the Switchboard corpus. Furthermore,\nby marginalizing over latent discourse relations at test time, we obtain a\ndiscourse informed language model, which improves over a strong LSTM baseline.","url_abs":"http://arxiv.org/abs/1603.01913v2","url_pdf":"http://arxiv.org/pdf/1603.01913v2.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":"a-latent-variable-recurrent-neural-network","repo_url":"https://github.com/jiyfeng/drlm","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"dialog-act-classification","task_name":"Dialog Act Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"implicit-discourse-relation-classification","task_name":"Implicit Discourse Relation Classification"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":null,"task_name":"Relation"},{"task_slug":"relation-classification","task_name":"Relation Classification"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1603.01913","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}