Baiyu Peng
Papers
1
Total Citations
23
H-Index
1
About
Baiyu Peng is a researcher advancing the frontier of safe reinforcement learning (RL), with a primary focus on developing algorithms that can operate reliably under uncertainty. His most cited work, "Model-Based Chance-Constrained Reinforcement Learning via Separated Proportional-Integral Lagrangian" (2022, 23 citations), tackles a critical challenge in real-world RL deployment: ensuring safety through probabilistic constraints. Peng introduces an innovative approach that separates proportional and integral control in the Lagrangian framework, enabling more stable and effective constraint satisfaction than traditional penalty or Lagrangian methods. This contribution is particularly significant for applications in robotics, autonomous systems, and industrial control, where safety violations can have severe consequences. By addressing the fundamental tension between exploration and safety in model-based RL, Peng's research provides a principled methodology for training agents that can learn optimal behaviors while respecting hard safety limits. His work represents an important step toward bridging the gap between theoretical RL advances and practical deployment in safety-critical environments, making him a notable emerging voice in the field of constrained reinforcement learning.
Research Focus
Key Achievements
Top Papers
- 1