Invalsi MATH, Invalsi ITA
收藏arXiv2024-03-27 更新2024-08-06 收录
下载链接：
http://arxiv.org/abs/2403.18697v1
下载链接
链接失效反馈官方服务：
资源简介：
Invalsi数据集由ISTI CNR创建，包含两个部分：Invalsi MATH和Invalsi ITA，分别用于评估语言模型在数学理解和意大利语语言理解方面的能力。这些数据集基于意大利学校系统中11至18岁学生的真实测试，包含422个问题，涉及多种题型如多选、真假判断、填空等。数据集的创建旨在为意大利语环境下的语言模型提供评估基准，解决现有模型在非英语环境中表现不佳的问题。
The Invalsi Dataset, developed by ISTI CNR, consists of two components: Invalsi MATH and Invalsi ITA, which are respectively designed to evaluate the capabilities of language models in mathematical comprehension and Italian language understanding. These datasets are sourced from real assessments administered to students aged 11 to 18 across the Italian school system, comprising a total of 422 questions covering various question formats including multiple-choice, true-false judgment, fill-in-the-blank, and more. The creation of this dataset aims to provide an evaluation benchmark for language models in the Italian-language context, addressing the underperformance of existing models in non-English environments.
提供机构：
ISTI CNR
创建时间：
2024-03-27
