Gave a talk on preference-robust fine-tuning of LLMs for public health at the NSF AI Institute for Societal Decision Making seminar.