Qiwei He

University of Science and Technology of China

Papers

Total Citations

H-Index

About

Qiwei He is a researcher in artificial intelligence, with a primary focus on deep reinforcement learning (DRL) and its application to complex, continuous control tasks. His major contributions center on advancing sample efficiency in environments with sparse rewards—a critical challenge in DRL. He is best known for developing "Soft Hindsight Experience Replay" (2020, 11 citations), which refines the classic Hindsight Experience Replay (HER) algorithm to enable more robust learning in robotic arm control and similar settings. Building on this, his work "Quantile Regression Hindsight Experience Replay" (2020, 3 citations) further integrates quantile regression to improve value estimation and policy stability. Though early in his career, He’s research addresses a fundamental bottleneck in DRL: making agents learn effectively from rare or delayed feedback. His methods have implications for robotics, autonomous systems, and any domain requiring efficient exploration. By tackling the brittleness of existing HER approaches, He is helping to pave the way for more reliable and practical reinforcement learning solutions.