Reinforcement Learning to Train Large Language Models to Explain Human Decisions arxiv.org 15 points by PaulHoule 18 hours ago