Jaekyung Cho
Papers
2
Total Citations
5
H-Index
2
About
Jaekyung Cho is a rising researcher in the field of reinforcement learning (RL) for robotics, specializing in making autonomous agents robust and reliable in the unpredictable real world. His primary research areas center on deep reinforcement learning, out-of-distribution (OOD) generalization, and self-supervised learning. Cho’s major contributions address a critical flaw in traditional RL: the tendency of agents to take overconfident, catastrophic actions when encountering unfamiliar states. In his 2022 work, "UNICON," he introduced an uncertainty-conditioned policy that dynamically adjusts behavior based on the agent’s confidence, ensuring safer navigation in novel scenarios. Building on this, his 2023 paper "SeRO" pioneered a self-supervised framework that allows robots to autonomously recover from OOD situations without human intervention. While his most-cited papers currently have 3 and 2 citations respectively, reflecting the early stage of his career, the conceptual novelty of his work is gaining traction. Cho’s research is particularly notable for its practical focus on bridging the simulation-to-reality gap, a persistent challenge in robotics. His innovative approach to uncertainty-aware learning positions him as a promising voice in the quest for truly autonomous, resilient robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2