Linran Zhao
The University of Texas at Austin
lrzhao@utexas.edu
Bio
Bo Zhao is a PhD student in computer science at the University of California, San Diego, advised by Professor Rose Yu. Her research focuses on symmetries in the parameter space of neural networks, as well as their impact on optimization, generalization, and the loss landscape. She received her M.S. in computer science from Georgia Institute of Technology and dual B.S. degrees in physics and computer science from the University of Illinois at Urbana-Champaign. Bo is a recipient of the DeepMind Fellowship and was a finalist for the Qualcomm Innovation Fellowship and the NVIDIA Graduate Fellowship.
Areas of Research
- Circuit Design
Integrated Circuit and System Technologies for Wireless Biomedical Devices
In many neural networks, different parameter values can yield the same loss, often due to underlying symmetries in the parameter space. We introduce a general framework for continuous symmetries based on equivariance in activation functions, revealing a new set of nonlinear, data-dependent symmetries. Using these symmetries, we derive topological properties of the minima and identify conserved quantities in gradient flows. As a practical application of parameter space symmetries, we present an algorithm called symmetry teleportation, which leverages parameter symmetries to search loss level sets for points with desired properties. This approach leads to improvements in both convergence and generalization.