LLMs average <5% on 2025 Math Olympiad; award each other 20x points
LLMs average <5% on 2025 Math Olympiad; award each other 20x points
arxiv.org
Proof or Bluff? Evaluating LLMs on 2025 USA Math Olympiad
Recent math benchmarks for large language models (LLMs) such as MathArena indicate that state-of-the-art reasoning models achieve impressive performance on mathematical competitions like AIME, with th...
"Notably, O3-MINI, despite being one of the best reasoning models, frequently skipped essential proof steps by labeling them as "trivial", even when their validity was crucial."