Sibo Gai

Papers

Total Citations

H-Index

About

Sibo Gai is a researcher at the forefront of efficient machine learning and embodied AI, with a focus on bridging the gap between powerful algorithms and real-world robotic systems. His work centers on two critical challenges: making deep learning models compact enough for edge devices, and enabling robots to learn continuously in dynamic environments. In his influential 2020 paper, "Knowledge Distillation for Model-Agnostic Meta-Learning" (15 citations), Gai pioneered a method to compress MAML-based few-shot learners, allowing them to run on resource-constrained hardware like mobile phones and small robots—a key step toward practical, on-device AI. More recently, his 2024 work, "Continual Reinforcement Learning for Quadruped Robot Locomotion" (8 citations), tackles the problem of catastrophic forgetting in legged robots. By developing a continual RL framework that balances plasticity for new tasks with stability for old ones, Gai enables quadruped robots to acquire locomotion skills sequentially without losing prior abilities. This research is vital for autonomous systems that must adapt to changing terrains and tasks over their lifetime. With a growing citation impact and a focus on deployable intelligence, Sibo Gai is shaping the future of adaptive, efficient, and lifelong learning machines.