Catherine R. Glossop

About

Catherine R. Glossop is a researcher focused on the robustness and reliability of reinforcement learning (RL) systems, particularly in continuous control domains. Her major contribution lies in systematically characterizing how state-of-the-art deep and robust RL algorithms perform under stress, using deliberate disturbance injection to expose vulnerabilities. In her most-cited work, "Characterising the Robustness of Reinforcement Learning for Continuous Control using Disturbance Injection" (2022, 2 citations), she leverages an open-source benchmark suite to conduct rigorous experiments that reveal critical failure modes in continuous action spaces. This research provides a foundational framework for evaluating and improving RL safety, directly impacting fields like robotics and autonomous systems where real-world deployment demands resilience. Glossop’s work is notable for bridging the gap between theoretical robustness and practical testing, offering a reproducible methodology that other researchers can adopt. Her findings are essential for students and engineers seeking to build more trustworthy AI systems, highlighting the importance of stress-testing algorithms beyond standard performance metrics.

Research Focus

Key Achievements

Top Papers

1. 1
   Characterising the Robustness of Reinforcement Learning for Continuous Control using Disturbance Injection

   2 citations · 2022

Key Collaborators

Contact & Links