Ye Pu

About

Ye Pu’s research lies at the intersection of reinforcement learning and next-generation communication networks, where he tackles fundamental challenges in data efficiency and real-time interactivity. His most influential work, “MBB: Model-Based Baseline for Efficient Reinforcement Learning” (2020), addresses a critical bottleneck in robotic control: while model-free RL excels in high-dimensional tasks, it suffers from severe data inefficiency. Pu’s model-based framework bridges this gap by integrating system dynamics into the learning process, enabling far more sample-efficient policy acquisition—a contribution that has garnered steady citations from the robotics and AI communities. More recently, Pu has ventured into the domain of haptic communication, as evidenced by his 2025 paper on explainable AI-assisted low-latency haptic feedback prediction for human-to-machine applications over passive optical networks. This work proposes an edge AI architecture that predicts haptic feedback from control signals, slashing latency to enable truly immersive teleoperation. By combining model-based RL principles with optical network optimization, Pu is pioneering a new class of intelligent, real-time control systems. His trajectory from foundational RL theory to applied networked intelligence marks him as a rising figure in both autonomous systems and communication engineering.

Research Focus

Key Achievements

Top Papers

1. 1
   MBB: Model-Based Baseline for Efficient Reinforcement Learning.

   3 citations · 2020
2. 2
   Explainable AI-assisted low-latency haptic feedback prediction for human-to-machine applications over passive optical networks

   1 citations · 2025

Key Collaborators

Contact & Links