Jakob Karalus
Papers
1
Total Citations
4
H-Index
1
About
Jakob Karalus is a researcher advancing the frontier of human-in-the-loop reinforcement learning (HRL), with a particular focus on making intelligent agents more accessible and interactive. His most-cited work, "Accelerating the Learning of TAMER with Counterfactual Explanations" (2022, 4 citations), addresses a critical bottleneck in HRL: the inefficiency of learning from human feedback alone. By integrating counterfactual explanations into the TAMER (Training an Agent Manually via Evaluative Reinforcement) framework, Karalus demonstrates how agents can not only learn from real-time human critiques but also understand *why* a different action would have been better. This approach dramatically accelerates the learning process, enabling even novice users to train service robots for new tasks naturally and interactively. His research bridges the gap between explainable AI and practical robot training, offering a pathway toward more intuitive human-robot collaboration. Karalus’s work is particularly significant for its potential to democratize AI training, allowing non-experts to shape agent behavior without deep technical knowledge. With a growing citation footprint, he is establishing himself as a key voice in making reinforcement learning more transparent, efficient, and user-friendly.
Research Focus
Key Achievements
Top Papers
- 1Accelerating the Learning of TAMER with Counterfactual Explanations4 citations · 2022