Grace Zhang

I am a fourth-year PhD student at USC, advised by Professor Gaurav Sukhatme. Prior to joining USC, I received my BS in EECS at UC Berkeley where I did research with Professor Sergey Levine .

Email / GitHub / Google Scholar

Research

I am interested in efficient robot learning methods for automating complex tasks through use of prior knowledge. Specifically, my research focuses on the generalization of learned policies to new environments, sample efficient reinforcement learning through multi-task learning, and the reuse of learned skills in hierarchical RL. My future work focuses on the use of multi-task data and general knowledge for more sample efficient task learning.

	Efficient Multi-Task Reinforcement Learning via Selective Behavior Sharing Grace Zhang, Ayush Jain, Injune Hwang, Shao-Hua Sun, Joseph J. Lim arXiv, 2023 paper / website /
	Policy Transfer across Visual and Dynamics Domain Gaps via Iterative Grounding Grace Zhang, Linghan Zhong, Youngwoon Lee, Joseph J. Lim Robotics Science and Systems, 2021 paper / website / code /
	MCP: Learning Composable Hierarchical Control with Multiplicative Compositional Policies Xue Bin Peng, Michael Chang, Grace Zhang, Pieter Abbeel, Sergey Levine Neural Information Processing Systems, 2019 paper / website /
	Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning Xue Bin Peng, Aviral Kumar, Grace Zhang, Sergey Levine arXiv Preprint, 2019 paper / website /