Online platforms that deliver ads and general recommendation systems are now integral to our daily routines. These platforms aim to maximize user engagement with their content, whether it involves ads or general recommendations. And yet, for all of their importance, researchers are still struggling to fully capture which factors drive user engagement on these platforms. Understanding these factors can not only unlock the possibility of increased revenue for the platforms, but can also help increase user satisfaction. Despite the multitude of models put forth to explain user engagement behavior, most focus on short-sighted/myopic preferences; i.e., users are assumed to make engagement decisions not caring about their prior interactions with the platform. In this paper, we take a different approach and aim to model users' evolving preferences.