Skip to content

Dataset · opencompass

AIME 2025 math olympiad problems

opencompass/AIME2025

Collection of English math problems from the AIME 2025 mathematical olympiad (rounds I and II) to evaluate LLM reasoning.

What's inside

English text dataset of less than 1 thousand records. It contains problems from the 2025 American Invitational Mathematics Examination (AIME) from the first and second rounds. The set is used as a standard benchmark to evaluate the mathematical capabilities of models.

What you can do with it

  • Benchmarking and verifying mathematical reasoning in reasoning models
  • Quality control after SFT or RL fine-tuning on mathematical data
Tasks
Question answering
Languages
English
Size
up to 1K rows
Downloads this month
11.7K
Likes
57

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("opencompass/AIME2025")