{"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/words-or-characters-fine-grained-gating-for","title":"Words or Characters? Fine-grained Gating for Reading Comprehension","arxiv_id":"1611.01724","date":"2016-11-06","proceeding":null,"authors":["Zhilin Yang","Bhuwan Dhingra","Ye Yuan","Junjie Hu","William W. Cohen","Ruslan Salakhutdinov"],"abstract":"Previous work combines word-level and character-level representations using\nconcatenation or scalar weighting, which is suboptimal for high-level tasks\nlike reading comprehension. We present a fine-grained gating mechanism to\ndynamically combine word-level and character-level representations based on\nproperties of the words. We also extend the idea of fine-grained gating to\nmodeling the interaction between questions and paragraphs for reading\ncomprehension. Experiments show that our approach can improve the performance\non reading comprehension tasks, achieving new state-of-the-art results on the\nChildren's Book Test dataset. To demonstrate the generality of our gating\nmechanism, we also show improved results on a social media tag prediction task.","url_abs":"http://arxiv.org/abs/1611.01724v2","url_pdf":"http://arxiv.org/pdf/1611.01724v2.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":"words-or-characters-fine-grained-gating-for","repo_url":"https://github.com/kimiyoung/fg-gating","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"reading-comprehension","task_name":"Reading Comprehension"},{"task_slug":"tag","task_name":"TAG"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/question-answering-on-squad11","task":"Question Answering","dataset":"SQuAD1.1","model":"Fine-Grained Gating","rank_in_archive_order":188,"of":213,"metrics":{"EM":"62.446","F1":"73.327"},"uses_additional_data":false},{"leaderboard":"/sota/question-answering-on-squad11-dev","task":"Question Answering","dataset":"SQuAD1.1 dev","model":"FG fine-grained gate","rank_in_archive_order":50,"of":55,"metrics":{"EM":"59.95","F1":"71.25"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1611.01724","atlas_url":"https://app.syntology.ai/?focus=1611.01724","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}