遇见数据集

pripro/imo-2025

收藏
Hugging Face2026-03-23 更新2026-03-29 收录
官方服务:

资源简介:

--- dataset_info: features: - name: id dtype: string - name: problem dtype: string - name: solution dtype: string splits: - name: train num_examples: 6 license: mit language: - en tags: - mathematics - olympiad - problem-solving - competition-math --- # IMO 2025 Problems Dataset This dataset contains the 6 problems from the 2025 International Mathematical Olympiad (IMO). The problems are formatted with proper LaTeX notation for mathematical expressions. ## Dataset Structure Each example contains: - `id`: Problem identifier (e.g., "2025-imo-p1") - `problem`: The problem statement with LaTeX mathematical notation - `solution`: The solution (currently set to null) ## Problem Types The dataset includes problems covering various mathematical areas: 1. **Problem 1**: Combinatorial geometry (sunny lines) 2. **Problem 2**: Euclidean geometry (circles and triangles) 3. **Problem 3**: Number theory (bonza functions) 4. **Problem 4**: Number theory (proper divisors and sequences) 5. **Problem 5**: Game theory (inekoalaty game) 6. **Problem 6**: Combinatorial geometry (grid tiling) ## Mathematical Notation Mathematical expressions are formatted using LaTeX: - Variables and expressions: `$x$`, `$n \geq 3$` - Display equations: `$$f(a) \text{ divides } b^a - f(b)^{f(a)}$$` - Sets: `$\mathbb{N}$`, `$\mathbb{R}$` - Special formatting: *sunny*, *bonza*, *proper divisor*, *inekoalaty game* ## Files - `imo_2025.json`: Full dataset in JSON format - `README.md`: This file ## Usage ```python from datasets import load_dataset # Load the dataset dataset = load_dataset("lmms-lab/imo-2025") # Access individual problems for problem in dataset['train']: print(f"Problem: {problem['id']}") print(f"Statement: {problem['problem']}") print() ``` Or load directly from JSON: ```python import json # Load from JSON with open("imo_2025.json", "r") as f: problems = json.load(f) # Access problems for problem in problems: print(f"ID: {problem['id']}") print(f"Problem: {problem['problem']}") print(f"Solution: {problem['solution']}") print() ``` ## Citation If you use this dataset in your research, please cite: ```bibtex @dataset{imo2025, title={IMO 2025 Problems Dataset}, author={LMMS Lab}, year={2025}, url={https://huggingface.co/datasets/lmms-lab/imo-2025} } ``` ## Source Problems are from the 2025 International Mathematical Olympiad. Original source: https://www.imo-official.org/problems.aspx ## License This dataset is released under the MIT License.

数据集信息: 特征: - 名称:id,数据类型:string - 名称:problem,数据类型:string - 名称:solution,数据类型:string 数据划分: - 名称:train,样本数量:6 许可证:MIT许可证 语言:英语(en) 标签: - 数学 - 奥林匹克竞赛 - 问题求解 - 竞赛数学 # 2025年国际数学奥林匹克(International Mathematical Olympiad,简称IMO)试题数据集 本数据集包含2025年国际数学奥林匹克的6道试题,所有题目均采用标准LaTeX语法格式化数学表达式。 ## 数据集结构 每个样本包含以下字段: - `id`:题目标识符(例如"2025-imo-p1") - `problem`:带有LaTeX数学符号的题目陈述 - `solution`:题目解答(当前设置为null) ## 试题类型 本数据集涵盖多个数学分支的试题: 1. **试题1**:组合几何(Combinatorial geometry),主题为阳光直线(sunny lines) 2. **试题2**:欧几里得几何(Euclidean geometry),主题为圆与三角形 3. **试题3**:数论(Number theory),主题为邦扎函数(bonza functions) 4. **试题4**:数论(Number theory),主题为真约数与序列 5. **试题5**:博弈论(Game theory),主题为伊科阿拉蒂博弈(inekoalaty game) 6. **试题6**:组合几何(Combinatorial geometry),主题为网格平铺(grid tiling) ## 数学符号规范 数学表达式采用LaTeX格式编写: - 变量与表达式:`$x$`、`$n \geq 3$` - 行间公式:`$$f(a) \text{ divides } b^a - f(b)^{f(a)}$$` - 集合符号:`$\mathbb{N}$`(自然数集)、`$\mathbb{R}$`(实数集) - 特殊格式标记:*阳光直线*、*邦扎函数*、*真约数*、*伊科阿拉蒂博弈* ## 文件说明 - `imo_2025.json`:JSON格式的完整数据集文件 - `README.md`:本说明文档 ## 使用方法 python from datasets import load_dataset # 加载数据集 dataset = load_dataset("lmms-lab/imo-2025") # 访问单个试题 for problem in dataset['train']: print(f"试题编号:{problem['id']}") print(f"题目陈述:{problem['problem']}") print() 或直接从JSON文件加载: python import json # 从JSON文件加载数据 with open("imo_2025.json", "r") as f: problems = json.load(f) # 访问试题 for problem in problems: print(f"试题编号:{problem['id']}") print(f"题目:{problem['problem']}") print(f"解答:{problem['solution']}") print() ## 引用方式 若您在研究中使用本数据集,请引用以下文献: bibtex @dataset{imo2025, title={IMO 2025 Problems Dataset}, author={LMMS Lab}, year={2025}, url={https://huggingface.co/datasets/lmms-lab/imo-2025} } ## 来源 本数据集的试题来源于2025年国际数学奥林匹克,原始来源链接:https://www.imo-official.org/problems.aspx ## 许可证 本数据集采用MIT许可证发布。

提供机构:
pripro
二维码
社区交流群
二维码
科研交流群
商业服务