Papers
5
Total Citations
93
H-Index
5
About
Jianyu Chen is a researcher at the forefront of safe reinforcement learning and autonomous robotic systems, whose work addresses some of the most pressing challenges in deploying intelligent agents in real-world environments. His most impactful contribution, "Model-Free Safe Reinforcement Learning Through Neural Barrier Certificate" (2023, 51 citations), introduced a principled framework for enforcing rigorous safety guarantees in RL without relying on overly conservative assumptions — a significant advance over prior methods that struggled to maintain reliable safety constraints. Complementing this, his work on uncertainty-aware reachability certificates extends safe RL into model-based settings, further reducing training-time violations critical for robotics applications. Beyond safety, Chen has made notable contributions to embodied AI, developing DoReMi, a system enabling large language models to detect and recover from execution failures in physical robot tasks. His research on decentralized motor skill learning and chance-constrained stochastic game solving further demonstrates his breadth across multi-robot coordination and locomotion. With nearly 100 cumulative citations across recent publications, Chen is establishing himself as a rising voice in the intersection of safe AI, robot learning, and language-grounded autonomy.
Research Focus
Key Achievements
Top Papers
- 1Model-Free Safe Reinforcement Learning Through Neural Barrier Certificate51 citations · 2023
- 2
- 3
- 4Decentralized Motor Skill Learning for Complex Robotic Systems6 citations · 2023
- 5Chance-Constrained Iterative Linear-Quadratic Stochastic Games5 citations · 2022