BenchmarkAIME 2025
AIME 2025
An AI reasoning evaluation benchmark composed of problems from the 2025 American Invitational Mathematics Examination (AIME). It is used to assess the ability of language models to perform reasoning on competition-level mathematics problems.
Timeline (last 90 days)
Sep 18
Using Reflective Recovery, DeepSeek-R1-Distill-Qwen-7B's accuracy on AIME 2025 improved from 30.0% to 37.5%.
Unverified50%