1 paper
Dalia Ali, Dora Zhao, Allison Koenecke +1
Although large language models (LLMs) are increasingly trained using human feedback for safety and alignment with human values, alignment decisions often overlook human social dive…