{"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/semantic-sentence-matching-with-densely","title":"Semantic Sentence Matching with Densely-connected Recurrent and Co-attentive Information","arxiv_id":"1805.11360","date":"2018-05-29","proceeding":null,"authors":["Seonhoon Kim","Inho Kang","Nojun Kwak"],"abstract":"Sentence matching is widely used in various natural language tasks such as\nnatural language inference, paraphrase identification, and question answering.\nFor these tasks, understanding logical and semantic relationship between two\nsentences is required but it is yet challenging. Although attention mechanism\nis useful to capture the semantic relationship and to properly align the\nelements of two sentences, previous methods of attention mechanism simply use a\nsummation operation which does not retain original features enough. Inspired by\nDenseNet, a densely connected convolutional network, we propose a\ndensely-connected co-attentive recurrent neural network, each layer of which\nuses concatenated information of attentive features as well as hidden features\nof all the preceding recurrent layers. It enables preserving the original and\nthe co-attentive feature information from the bottommost word embedding layer\nto the uppermost recurrent layer. To alleviate the problem of an\never-increasing size of feature vectors due to dense concatenation operations,\nwe also propose to use an autoencoder after dense concatenation. We evaluate\nour proposed architecture on highly competitive benchmark datasets related to\nsentence matching. Experimental results show that our architecture, which\nretains recurrent and attentive features, achieves state-of-the-art\nperformances for most of the tasks.","url_abs":"http://arxiv.org/abs/1805.11360v2","url_pdf":"http://arxiv.org/pdf/1805.11360v2.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":[],"tasks":[{"task_slug":"natural-language-inference","task_name":"Natural Language Inference"},{"task_slug":"paraphrase-identification","task_name":"Paraphrase Identification"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"sentence","task_name":"Sentence"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/natural-language-inference-on-snli","task":"Natural Language Inference","dataset":"SNLI","model":"Densely-Connected Recurrent and Co-Attentive Network Ensemble","rank_in_archive_order":13,"of":98,"metrics":{"% Test Accuracy":"90.1","% Train Accuracy":"95.0","Parameters":"53.3m"},"uses_additional_data":false},{"leaderboard":"/sota/natural-language-inference-on-snli","task":"Natural Language Inference","dataset":"SNLI","model":"Densely-Connected Recurrent and Co-Attentive Network","rank_in_archive_order":25,"of":98,"metrics":{"% Test Accuracy":"88.9","% Train Accuracy":"93.1","Parameters":"6.7m"},"uses_additional_data":false},{"leaderboard":"/sota/natural-language-inference-on-snli","task":"Natural Language Inference","dataset":"SNLI","model":"Densely-Connected Recurrent and Co-Attentive Network (encoder)","rank_in_archive_order":55,"of":98,"metrics":{"% Test Accuracy":"86.5","% Train Accuracy":"91.4","Parameters":"5.6m"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1805.11360","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}