Bio
I’m Michael Ito, a Ph.D. student in Computer Science and Engineering at the University of Michigan. I am advised by Jenna Wiens and also work closely with Danai Koutra. I am supported by the Department of Energy Computational Science Graduate Fellowship (DOE CSGF).
Research
My research spans graph learning, equivariant learning, and drug discovery. I study how existing equivariant and symmetry-aware architectures on graphs trade off expressive power, generalization, and computational efficiency, particularly in molecular and protein modeling. Guided by these insights, I develop new ways of capturing symmetry and structure, such as probabilistic symmetry-breaking approaches, new canonicalization schemes, and architectural invariances tailored to molecular and protein graphs, that are both theoretically grounded and empirically effective on challenging drug discovery tasks.