Researchers Develop Method to Train LLMs for High-Performance Chess with Accurate Move Explanations