Zhongjun Zhang



Taking the scenic route

In reinforcement learning, we balance exploration and exploitation. In life, I let exploration take the wheel. I love road trips, unfamiliar places, and the occasional scenic detour that turns into the highlight of the trip:)