Unsupervised pre-training has recently become the bedrock for computer vision and natural
language processing. In reinforcement learning (RL), goal-conditioned RL can potentially …

Offline reinforcement learning typically introduces a hierarchical structure to solve the long-
horizon problem so as to address its thorny issue of variance accumulation. Problems of …

Learning contact-rich manipulation skills is essential. Such skills require the robots to
interact with the environment with feasible manipulation trajectories and suitable compliance …

Deep generative models (DGMs) have demonstrated great success across various domains,
particularly in generating texts, images, and videos using models trained from offline data …

Diffusion-based generative methods have proven effective in modeling trajectories with
offline datasets. However, they often face computational challenges and can falter in …

Developing agents that can execute multiple skills by learning from pre-collected datasets is
an important problem in robotics, where online interaction with the environment is extremely …

Autonomous driving (AD) holds the potential to revolutionize transportation efficiency, but its
success hinges on robust behavior planning (BP) mechanisms. Reinforcement learning (RL) …

The ability to plan actions on multiple levels of abstraction enables intelligent agents to solve
complex tasks effectively. However, learning the models for both low and high-level …

Pre-training on task-agnostic large datasets is a promising approach for enhancing the
sample efficiency of reinforcement learning (RL) in solving complex tasks. We present …

Learning-based vehicle planning is receiving increasing attention with the emergence of
diverse driving simulators and large-scale driving datasets. While offline reinforcement …