2 ms·Enigmata: Scaling Logical Reasoning In LLMs With Synthetic Verifiable Puzzles2 points by optimalsolver 1y ago