Romanian Informatics Olympiad Challenges

该数据集包含了304个来自罗马尼亚信息学奥林匹克竞赛的挑战题目，时间跨度从2002年至2023年。它专注于大型语言模型（如LLM）使用C++和Python编写的解决方案，并特别包括了OJI问题，覆盖了不同年级的挑战。此外，该数据集还评估了大型语言模型，如GPT-4、Gemini和CodeStral在竞技编程环境中的表现。规模上，数据集共有304个挑战题目，其任务是对大型语言模型在解决竞技编程问题时的性能进行评估。

This dataset contains 304 competitive programming problems sourced from the Romanian Informatics Olympiad, spanning the period from 2002 to 2023. It focuses on solutions written in C++ and Python by large language models (LLMs), and specifically includes OJI problems covering challenges for different grade levels. Additionally, this dataset evaluates the performance of prominent large language models including GPT-4, Gemini, and CodeStral in competitive programming environments. The primary goal of this dataset is to assess the performance of large language models when solving competitive programming problems.