Longbo Huang

Provable Generalization of Overparameterized Meta-learning Trained with SGD

2022 · arXiv (Cornell University)

Despite the superior empirical success of deep meta-learning, theoretical understanding of overparameterized meta-learning is still limited. This paper studies the generalization of a widely used meta-learning approach, Model-Agnostic Meta-Learning (MAML), which aims to find a …

Improving Generalization and Data Efficiency with Diffusion in Offline Multi-agent RL

2023 · arXiv (Cornell University)

We present a novel Diffusion Offline Multi-agent Model (DOM2) for offline Multi-Agent Reinforcement Learning (MARL). Different from existing algorithms that rely mainly on conservatism in policy design, DOM2 enhances policy expressiveness and diversity based on …

One is More: Diverse Perspectives within a Single Network for Efficient DRL

2023 · arXiv (Cornell University)

Deep reinforcement learning has achieved remarkable performance in various domains by leveraging deep neural networks for approximating value functions and policies. However, using neural networks to approximate value functions or policy functions still faces challenges, …