Jeongwoo Lee
Papers
1
Total Citations
2
H-Index
1
About
Jeongwoo Lee is a rising researcher in artificial intelligence, with a focus on cross-domain policy adaptation and semantic skill learning. Their most cited work, "SemTra: A Semantic Skill Translator for Cross-Domain Zero-Shot Policy Adaptation" (2024, 2 citations), introduces a novel framework for enabling AI agents to transfer learned behaviors across different domains without additional training. Lee's key contribution lies in developing semantically interpretable "skills"—expert behavior patterns that can be translated between environments using interleaved multi-modal user inputs. This approach allows agents to tackle novel, long-horizon tasks by leveraging prior knowledge, significantly advancing zero-shot generalization in robotics and embodied AI. Though early in their career, Lee's work addresses a critical bottleneck in AI: the ability to adapt to unseen scenarios without costly retraining. Their research bridges semantic reasoning and reinforcement learning, offering a pathway toward more flexible and human-interpretable autonomous systems. With growing interest in foundation models and transfer learning, Lee's contributions are poised to influence future work in cross-domain policy learning and human-AI collaboration.
Research Focus
Key Achievements
Top Papers
- 1