Keqiang Li
Papers
1
Total Citations
53
H-Index
1
About
Keqiang Li is a rising force in reinforcement learning (RL), with a focused expertise in developing more stable and sample-efficient algorithms for complex decision-making and control tasks. His most-cited work, "Distributional Soft Actor-Critic With Three Refinements" (2025, 53 citations), directly tackles a critical bottleneck in model-free RL: the performance degradation caused by inaccurate value estimation, particularly the overestimation of Q-values. By introducing three key refinements to the Soft Actor-Critic framework, Li’s contribution offers a practical pathway to more robust policy learning, addressing a fundamental challenge that has long limited the real-world deployment of RL. This work highlights his ability to identify and solve core theoretical issues with tangible algorithmic improvements. As an emerging scholar, Li’s research is carving a niche in making RL not just powerful, but also reliable, positioning him as a promising contributor to the next generation of intelligent control systems.
Research Focus
Key Achievements
Top Papers
- 1Distributional Soft Actor-Critic With Three Refinements53 citations · 2025