Dataset · yentinglin
AIME 2025 math olympiad problems
yentinglin/aime_2025
A set of 30 math competition problems from AIME I and AIME II 2025 for evaluating model reasoning.
What's inside
The dataset collects 30 math problems from the 2025 AIME competitions: 15 problems each from AIME I and AIME II stages. The data is presented in text and tabular formats. The dataset has a small volume and is used to test models' mathematical capabilities.
What you can do with it
- Benchmarking mathematical and logical reasoning of large language models (LLMs)
- Evaluating the quality of reasoning models on recent competition problems
- Size
- up to 1K rows
- Downloads this month
- 27.1K
- Likes
- 12
How to load it
from datasets import load_dataset
ds = load_dataset("yentinglin/aime_2025")