I have broad interests in developing artificial intelligence, with a primary interest in the role of language, as a vehicle of thought and communication, in both artificial and human intelligence. These days, I'm particularly fond of foundation models, including both large language models (LLMs) and multimodal ones, and I believe they are an important step towards more general AI and should be studied thoroughly. I enjoy developing language agents that, in a robust and trustworthy way, automate a lot of our tedious tasks and make the computing world more accessible to all. I like to make things that work, which often means operating at scale and in real-world settings. For that I also deeply care about generalizability, interpretability, efficiency, and robustness. Finally, I believe in translational research and seek to leverage AI capabilities to empower a variety of domains such as biology and biomedicine, and meanwhile seek inspirations from these disciplines to develop better AI.