Paula Gradu
Papers
1
Total Citations
5
H-Index
1
About
Paula Gradu is a researcher advancing the frontier of differentiable control and robotics, with a focus on bridging simulation and real-world learning. Her key research areas include differentiable physics engines, gradient-based control, and benchmarking for autonomous systems. Gradu’s most notable contribution is the development of **Deluca**, an open-source differentiable control library that provides natively differentiable physics and robotics environments. This work enables auto-differentiation through simulation dynamics, allowing for rapid training of control policies using gradient-based methods—a significant departure from traditional model-free reinforcement learning. The accompanying benchmarking suite offers standardized environments and methods, facilitating reproducible comparisons across the field. While her most-cited paper has garnered 5 citations, its impact lies in laying foundational infrastructure for a growing community of researchers exploring differentiable simulations for robot learning. Gradu’s work is particularly valuable for students and engineers seeking efficient, principled approaches to control in continuous domains, and her library promises to accelerate progress in areas from manipulation to locomotion by making gradient information directly accessible through the dynamics of complex systems.
Research Focus
Key Achievements
Top Papers
- 1