Guoqing Liu is a Senior Researcher at
Microsoft Research AI for Science based in Cambridge, UK. His research focuses on Reinforcement Learning (RL), Large Language Models (LLMs), and AI Co-Scientists. He develops LLMs post-training methods and autonomous RL agents to accelerate chemistry and drug discovery, with representative works including
QFANG,
NatureLM,
PDVN, and
the TDPO series.
Previously, he was a Senior Researcher and a joint PhD student at
Microsoft Research Asia, where he worked on (deep) reinforcement learning, particularly state abstraction and representation, policy gradient methods, and sample efficiency. Representative projects include
Suphx: The World's Best Mahjong AI and
Inspector: Automated Game Testing with Xbox Studios.
He received his Ph.D. from the
University of Science and Technology of China through a joint program with
Microsoft Research Asia (2016-2021), supervised by
Tie-Yan Liu and
Nenghai Yu.
