Hi! I’m Yixin Dong, a Ph.D. student at Carnegie Mellon University, fortunate to be advised by Prof. Tianqi Chen. I hold a Bachelor of Science in Computer Science and Technology from the ACM Honors Class at Shanghai Jiao Tong University.

I am building efficient and verifiable AI agents. I regard it as a crucial pathway to AGI.

I am also interested in ML systems and ML compilers. I am a dedicated advocate of open source, leading several well-adopted projects, including MLC-LLM and Apache TVM Unity, where I serve as an Apache TVM Reviewer. I had the privilege of spending a wonderful time at the University of Washington, working with Prof. Luis Ceze.

We have been driving the XGrammar project forward recently. It is a zero-overhead LLM constrained generation engine and has already been adopted by several well-known LLM frameworks.

Feel free to explore my CV to learn more. I’m always open to collaborating with passionate, like-minded individuals.

I am open to advising CMU undergrads and master’s students with backgrounds or interests in LLM post-training (especially RL), CUDA, or (ML) compilers. I’m also happy to chat about research and PhD applications with students from underrepresented groups or anyone in need. Please reach out by email.

Beyond research, I’m a cinephile and a lover of contemporary dance. My favorite movies are Hiroshima mon amour and Il buono, il brutto, il cattivo. If you share similar interests, let’s grab a coffee and chat!