About Me

I'm an incoming assistant professor at TTIC, and a current faculty fellow at NYU CDS. Previously, I received my PhD from Berkeley EECS, where I was part of the Berkeley NLP Group and advised by Dan Klein. Before that, I was an undergrad at Brown University, where I majored in math and linguistics and was advised by Ellie Pavlick.

Research

Language can be seen as the solution to a multi-agent collaborative game. Under this view, there are many natural questions to ask:

• How do properties of our environment shape the language we use? (Tomlin, et al. 2025)
• How can we communicate efficiently under cognitive resource constraints? (Lin, et al. 2024)
• Can we design training objectives to incentivize honest and helpful language use? (Liao, et al. 2024)

My research uses tools from natural language processing and multi-agent reinforcement learning to answer such questions, in order to improve the capabilities and safety of future AI systems. Recently, I've been thinking about how to design objectives to explain superhuman AI behavior in a human-interpretable way, and how to reduce reward hacking and deception in LLM training.

I'm broadly interested in computational cognitive science, linguistics, reinforcement learning, human-computer interaction, and AI safety. I have worked on a variety of other topics, including computer crossword solving, protolanguage reconstruction, and computational psycholinguistics. To learn more about my research interests, please check out my papers or my blog.

Lab

I will be running a research lab at TTIC focused on natural language processing, interaction, and reinforcement learning. To learn more about my group and current students, please check out my group page.