Vilde B. Gjarum
Papers
1
Total Citations
12
H-Index
1
About
Vilde B. Gjarum is a researcher at the intersection of reinforcement learning and explainable artificial intelligence, with a focus on making autonomous systems both effective and interpretable. Her work addresses a critical challenge in modern robotics: while deep reinforcement learning produces powerful control agents, their black-box nature hinders deployment in high-stakes environments. Gjarum’s most cited paper, "Approximating a deep reinforcement learning docking agent using linear model trees" (2021, 12 citations), introduces a novel method to distill a complex deep RL policy into a transparent, rule-based model. By using linear model trees, she demonstrates how to preserve high performance while enabling human-understandable decision-making—a vital step for safety-critical applications like autonomous docking. This contribution bridges the gap between cutting-edge learning algorithms and practical engineering requirements, offering a pathway toward trustworthy AI in robotics. Gjarum’s research is particularly impactful for students and engineers seeking to deploy reinforcement learning in real-world systems where interpretability is non-negotiable, such as in maritime or aerospace operations. Her work exemplifies a rigorous, application-driven approach to advancing both the science and the safety of autonomous agents.
Research Focus
Key Achievements
Top Papers
- 1