[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-272650-105":59,"doc-detail-272650-en":131},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","race-large-scale-reading-comprehension-dataset-from-examinations","RACE - Large-Scale Reading Comprehension Dataset From Examinations","","RACE is introduced as a new benchmark dataset for reading comprehension evaluation. Built from English exams written for middle- and high-school Chinese students aged 12–18, it contains nearly 28,000 passages and about 100,000 questions. Questions require substantially more reasoning than other benchmarks, creating a larger gap between state-of-the-art performance (43%) and human ceiling (95%). The dataset supports research and assessment for machine comprehension, with public availability of data and code.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/race-large-scale-reading-comprehension-dataset-from-examinations/272650/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/race-large-scale-reading-comprehension-dataset-from-examinations/272650.png","ImageObject",300,407,{"name":92,"@type":93},"Seraphina","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-04","2026-09-15",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",11,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"What is RACE designed to evaluate?","Question",{"text":113,"@type":114},"RACE is a benchmark dataset for evaluating methods on the reading comprehension task, focusing on a model’s understanding and reasoning about text.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"How was RACE constructed and what does it contain?",{"text":118,"@type":114},"RACE was collected from English exams for Chinese middle- and high-school students (ages 12–18) and includes nearly 28,000 passages and around 100,000 questions.",{"name":120,"@type":111,"acceptedAnswer":121},"Why are RACE questions considered harder than those in other benchmarks?",{"text":122,"@type":114},"RACE places a much larger proportion of questions that require reasoning, and its difficulty leads to a significant performance gap between state-of-the-art models and human experts.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},272650,1789466690,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":52,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":130,"read_time":144,"preview_page_count":4},962075114101,"https://ap-avatar.wpscdn.com/avatar/e000253a75eb197efd?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780044092746381165","RACE: Large-scale ReAding Comprehension Dataset From Examinations  \nGuokun Lai⇤ and Qizhe Xie⇤ and Hanxiao Liu and Yiming Yang and Eduard Hovy {guokun, qzxie, hanxiaol, yiming, [hovy](hovy}@cs.cmu.edu)[}](hovy}@cs.cmu.edu)[@cs.cmu.edu](hovy}@cs.cmu.edu)[ ](hovy}@cs.cmu.edu)Language Technologies Institute  \nCarnegie Mellon University  \nPittsburgh, PA 15213  \nAbstract  \nWe present RACE, a new dataset for benchmark evaluation of methods in the reading comprehension task. Collected from the English exams for middle and high school Chinese students in the age range between 12 to 18, RACE consists of near 28,000 passages and near  \n100,000 questions generated by human experts (English instructors), and covers a variety of topics which are carefully designed for evaluating the students’ability in understanding and reasoning. In particular, the proportion of questions that requires reasoning is much larger in RACE than that in other benchmark datasets for reading comprehension, and there is a significant gap between the performance of the state-of-the-art models (43%) and the ceiling human performance (95%) . We hope this new dataset can serve as a valuable resource for research and evaluation in machine comprehension. The dataset is freely available at [http://www.cs.cmu.edu/](http://www.cs.cmu.edu/)˜glai1/data/race/ and the code is available at [https://github.com/](https://github.com/)[ ](https://github.com/)qizhex/RACE_AR_baselines  \n1 Introduction  \nConstructing an intelligence agent capable of understanding text as people is the major challenge of NLP research. With recent advances in deep learning techniques, it seems possible to achieve human-level performance in certain language understanding tasks, and a surge of effort has been devoted to the machine comprehension task where people aim to construct a system with the ability to  \n indicates equal contribution  \nanswer questions related to a document that it has to comprehend (Chen et al., 2016 ; Kadlec et al., 2016 ; Dhingra et al., 2016 ; Yang et al., 2017) .  \nTowards this goal, several large-scale datasets (Rajpurkar et al., 2016 ; Onishi et al., 2016 ; Hill et al., 2015 ; Trischler et al., 2016 ; Hermann et al., 2015) have been proposed, which allow researchers to train deep learning systems and obtain results comparable to the human performance. While having a suitable dataset is crucial for evaluating the system’s true ability in reading comprehension, the existing datasets suffer several critical limitations. Firstly, in all datasets, the candidate options are directly extracted from the context (asa single entity or a text span), which leads to the fact that lots of questions can be solved trivially via word-based search and context-matching without deeper reasoning; this constrains the types of questions as well. Secondly, answers and questions of most datasets are either crowd-sourced or automatically-generated, bringing a significant amount of noises in the datasets and limits the ceiling performance by domain experts, such as 82% for Childrens Book Test and 84% for Who-didWhat. Yet another issue in existing datasets is that the topic coverages are often biased due to the specific ways that the data were initially collected, making it hard to evaluate the ability of systems in text comprehension over a broader range of topics. To address the aforementioned limitations, we constructed a new dataset by collecting a large set of questions, answers and associated passages in the English exams for middle-school and high-school Chinese students within the 12Ð18 age range. Those exams were designed by domain experts (instructors) for evaluating the reading comprehension ability of students, with ensured quality and broad topic coverage. Furthermore, the answers by machines or by humans can be objectively graded for evaluation  \n785  \nProceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 785–794 Copenhagen, Denmark, September 7–11, 201","cbCaiguVlES4O93N","https://ap.wps.com/l/cbCaiguVlES4O93N","pdf",537659,"English","# Abstract\n# Introduction","[{\"question\":\"What is RACE designed to evaluate?\",\"answer\":\"RACE is a benchmark dataset for evaluating methods on the reading comprehension task, focusing on a model’s understanding and reasoning about text.\"},{\"question\":\"How was RACE constructed and what does it contain?\",\"answer\":\"RACE was collected from English exams for Chinese middle- and high-school students (ages 12–18) and includes nearly 28,000 passages and around 100,000 questions.\"},{\"question\":\"Why are RACE questions considered harder than those in other benchmarks?\",\"answer\":\"RACE places a much larger proportion of questions that require reasoning, and its difficulty leads to a significant performance gap between state-of-the-art models and human experts.\"}]","RACE - Large-Scale Reading Comprehension Dataset From Examinations | PDF",25]