Yiqi "Andrew" Liu

I am Yiqi Liu (刘逸奇), and I usually go by Andrew. I am a physics Ph.D. student at Princeton University advised by Jo Dunkley. My research connects understanding our own Galaxy with learning how the Universe began. With the Simons Observatory, I study the cosmic microwave background (CMB), light left over from the early Universe, to search for evidence of primordial gravitational waves.

Galactic dust produces a much brighter polarized signal than the one we seek, and errors in modeling it can mimic a cosmological signal. I use controlled simulations to understand when foreground models fail. I am developing template-fitting tools to measure how dust emission changes across frequencies in Planck and Simons Observatory maps, and extending these diagnostics to probe different angular scales.

My broader interests include scientific AI and reliable statistical inference. I am exploring self-supervised learning and simulation-based inference to extract information about dust structure that conventional power spectra can miss, and to cross-check cosmological conclusions under different modeling assumptions. I also contribute to Terminal-Bench-Science, designing and reviewing tasks that evaluate AI agents’ scientific reasoning.

Before Princeton, I studied applied mathematics and statistics, physics, and mathematics at Johns Hopkins University, with a minor in computer science. There I worked with Tobias Marriage and Charles Bennett on the CLASS experiment, and with Nadia Zakamska and Hsiang-Chih Hwang on stellar astrophysics.