Vincent Duplessis
Papers
1
Total Citations
7
H-Index
1
About
Vincent Duplessis is a researcher at the forefront of reinforcement learning (RL) and its application to autonomous robotic systems. His work centers on the critical challenge of reward function engineering—the process of defining the goals that guide an RL agent’s behavior. His most cited paper, "MoVEMo: A Structured Approach for Engineering Reward Functions" (2018, 7 citations), introduces a formal methodology for designing and verifying reward structures, moving beyond ad-hoc trial-and-error to ensure that agents learn safe, efficient, and intended behaviors. This contribution is foundational for deploying RL in real-world robotics, where poorly specified rewards can lead to catastrophic outcomes. By providing a systematic framework, Duplessis helps bridge the gap between theoretical RL and practical, reliable autonomy. His work is essential reading for students and researchers seeking to build robust, goal-driven AI systems that can operate with predictable success in complex environments.
Research Focus
Key Achievements
Top Papers
- 1MoVEMo: A Structured Approach for Engineering Reward Functions7 citations · 2018