{"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/direct-output-connection-for-a-high-rank","title":"Direct Output Connection for a High-Rank Language Model","arxiv_id":"1808.10143","date":"2018-08-30","proceeding":"EMNLP 2018 10","authors":["Sho Takase","Jun Suzuki","Masaaki Nagata"],"abstract":"This paper proposes a state-of-the-art recurrent neural network (RNN)\nlanguage model that combines probability distributions computed not only from a\nfinal RNN layer but also from middle layers. Our proposed method raises the\nexpressive power of a language model based on the matrix factorization\ninterpretation of language modeling introduced by Yang et al. (2018). The\nproposed method improves the current state-of-the-art language model and\nachieves the best score on the Penn Treebank and WikiText-2, which are the\nstandard benchmark datasets. Moreover, we indicate our proposed method\ncontributes to two application tasks: machine translation and headline\ngeneration. Our code is publicly available at:\nhttps://github.com/nttcslab-nlp/doc_lm.","url_abs":"http://arxiv.org/abs/1808.10143v2","url_pdf":"http://arxiv.org/pdf/1808.10143v2.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":"direct-output-connection-for-a-high-rank","repo_url":"https://github.com/nttcslab-nlp/doc_lm","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"constituency-parsing","task_name":"Constituency Parsing"},{"task_slug":"headline-generation","task_name":"Headline Generation"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"translation","task_name":"Translation"},{"task_slug":"high","task_name":"Vocal Bursts Intensity Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/constituency-parsing-on-penn-treebank","task":"Constituency Parsing","dataset":"Penn Treebank","model":"LSTM Encoder-Decoder + LSTM-LM","rank_in_archive_order":17,"of":27,"metrics":{"F1 score":"94.47"},"uses_additional_data":false},{"leaderboard":"/sota/language-modelling-on-penn-treebank-word","task":"Language Modelling","dataset":"Penn Treebank (Word Level)","model":"AWD-LSTM-DOC x5","rank_in_archive_order":8,"of":43,"metrics":{"Params":"185M","Test perplexity":"47.17","Validation perplexity":"48.63"},"uses_additional_data":false},{"leaderboard":"/sota/language-modelling-on-penn-treebank-word","task":"Language Modelling","dataset":"Penn Treebank (Word Level)","model":"AWD-LSTM-DOC","rank_in_archive_order":16,"of":43,"metrics":{"Params":"23M","Test perplexity":"52.38","Validation perplexity":"54.12"},"uses_additional_data":false},{"leaderboard":"/sota/language-modelling-on-wikitext-2","task":"Language Modelling","dataset":"WikiText-2","model":"AWD-LSTM-DOC x5","rank_in_archive_order":20,"of":38,"metrics":{"Number of params":"185M","Test perplexity":"53.09","Validation perplexity":"54.19"},"uses_additional_data":false},{"leaderboard":"/sota/language-modelling-on-wikitext-2","task":"Language Modelling","dataset":"WikiText-2","model":"AWD-LSTM-DOC","rank_in_archive_order":24,"of":38,"metrics":{"Number of params":"37M","Test perplexity":"58.03","Validation perplexity":"60.29"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1808.10143","atlas_url":"https://app.syntology.ai/?focus=1808.10143","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}