Can Large Language Models Change User Preference Adversarially?

Varshini Subhash

Pretrained large language models (LLMs) are becoming increasingly powerful and ubiquitous in mainstream applications such as being a personal assistant, a dialogue model, etc. As t…