Another 9 open Erdos problems solved, this time by DeepMind team. Interesting loop of LLM - Lean agents working autonomously, and only after it's verified formally, going through human review.RedditMay 24, 2026 10:45ShareThis tweet cannot be embedded.Open direct link