Sungjae Nah

About

Sungjae Nah is a researcher at the intersection of reinforcement learning and probabilistic modeling, with a focus on enabling robust decision-making in complex, noisy physical systems. His work addresses fundamental challenges in robotics and autonomous systems, particularly how agents can learn and adapt under uncertainty. In his most-cited paper, "Distributional and hierarchical reinforcement learning for physical systems with noisy state observations and exogenous perturbations" (2023, 10 citations), Nah introduces a novel framework that combines distributional and hierarchical RL to handle both sensory noise and unpredictable environmental disturbances—a critical step toward deploying RL in real-world settings. Earlier, his work on "Multi-output Infinite Horizon Gaussian Processes" (2021) advanced online learning from noisy sensory streams, enabling robots to continuously model and predict uncertain dynamical environments. By bridging theory and application, Nah’s contributions are shaping how machines learn from noisy, high-dimensional data, with implications for fields from autonomous navigation to industrial control. His growing citation record reflects the practical relevance of his research, positioning him as an emerging voice in the quest for more resilient and adaptive intelligent systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Distributional and hierarchical reinforcement learning for physical systems with noisy state observations and exogenous perturbations

   10 citations · 2023
2. 2
   Multi-output Infinite Horizon Gaussian Processes

   2 citations · 2021

Key Collaborators

Contact & Links