Adding one irrelevant sentence to math problems causes AI systems to make confident mistakes over 300 percent more.

A few months before the 2025 International Mathematical Olympiad (IMO) in July, a three-person team at OpenAI made a long bet that they could use the competition’s brutally tough problems to train an ...

Artificial intelligence has moved from checking homework to attacking problems that professional mathematicians once treated as out of reach. Systems tuned for symbolic reasoning are now cracking long ...