Finetuning LLMs for Human Behavior Prediction in Social Science Experiments Paper • 2509.05830 • Published Sep 6