Six AI systems achieved perfect scores on IMO-level mathematics problems in 2025, utilizing Lean-verified proofs. This breakthrough was enabled by a highly stringent verification process, which ensures absolute correctness in the AI's solutions. The success highlights the potential of reinforcement learning from human feedback (RLHF) when paired with rigorous verification methods. AI
Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →
IMPACT Demonstrates AI's capability for perfect mathematical reasoning, potentially impacting formal verification and theorem proving.
RANK_REASON AI systems achieved perfect scores on a formal mathematics competition using verified proofs.