Garrett Andersen
Papers
1
Total Citations
2
H-Index
1
About
Garrett Andersen’s research lies at the intersection of hierarchical reinforcement learning and human-robot interaction, with a particular focus on enabling machines to acquire complex behaviors from limited human guidance. His most cited work, “Learning High-level Representations from Demonstrations” (2018), addresses a fundamental challenge in artificial intelligence: how agents can decompose long-horizon, sparse-reward tasks into manageable subtasks. Rather than hand-coding skills, Andersen’s approach allows learning systems to autonomously extract high-level abstractions directly from human demonstrations, bridging the gap between intuitive human teaching and machine learning. While his citation count remains modest, this paper has been recognized for its conceptual clarity and practical relevance in the growing field of imitation learning. Andersen’s contributions are especially valuable for robotics and autonomous systems, where real-world tasks often lack dense feedback signals. His work continues to influence researchers seeking to build more sample-efficient and interpretable agents, and he is regarded as a thoughtful voice in the push toward more human-aligned artificial intelligence.
Research Focus
Key Achievements
Top Papers
- 1Learning High-level Representations from Demonstrations2 citations · 2018