Osvaldo Quinjica

I’m a PhD student at University of Maryland–College Park, working on how we can build AI systems that reason reliably and act in line with what people actually want.

My research sits at the intersection of reinforcement learning and alignment. My research focuses on logical reasoning over learned concepts, exploring how AI systems can infer, generalize, and reason about rules from learned representations. Alongside this, I study multi-turn RL: training agents that stay coherent and goal-oriented over long interactions, where feedback is delayed and a single misstep can compound.

I also work on multilingual AI, with the goal of making AI systems more accessible across languages.

Before my PhD, I did research at Apple in Beijing, and at the Institute of Computing Technology, Chinese Academy of Sciences.