1 paper
Priyadarsi Mishra, Arjun Beniwal, Ziyu Yi +2
Large language model (LLM) alignment often assumes that the user interacting with a model is also the primary beneficiary of its behavior. Yet in many consequential settings, one p…