Publications

We propose a new radiology report generation method that aligns the pre-trained model with multiple human preferences via preference-guided multi-objective optimization reinforcement learning.

We propose Forward KL regularized Preference optimization for aligning Diffusion policies to align the diffusion policy with preferences, learning to align the policy output with human intents in various tasks.

We propose a novel dynamic policy constraint that restricts the learned policy on the samples generated by the exponentional moving average of previously learned policies for offline RL.

We propose a novel representation-based approach to measure the domain gap, where the representation is learned through a contrastive objective by sampling transitions from different domains.

We propose SAM-E, a novel architecture for robot manipulation by leveraging a vision-foundation model for generalizable scene understanding and sequence imitation for long-term action reasoning.

We propose a novel unsupervised RL framework via an ensemble of skills, where each skill performs partition exploration based on the state prototypes.

We adopt multi-task preferences as a unified condition for both single- and multi-task decision-making, and propose preference representations aligned with preference labels.

We construct an example to show the information-theoretical improvement in sample efficiency achieved by goal relabeling and develop an RL algorithm called GOALIVE.

We consider dynamics adaptation settings where there exists dynamics mismatch between the source domain and the target domain, and one can get access to sufficient source domain data, while can only have limited interactions with the target domain.

We consider a novel risk-sensitive perspective to enhance the robustness of legged locomotion.

We propose Optimistic Value Distribution Explorer (OVD-Explorer) to achieve a noise-aware optimistic exploration for continuous control.

We introduce a novel framework that leverages a unified discrete diffusion to combine generative pre-training on human videos and policy fine-tuning on a small number of action-labeled robot videos.

We aim to investigate the effectiveness of a single diffusion model in modeling large-scale multi-task offline data, which can be challenging due to diverse and multimodal data distribution.

We reveal the limitations of these methods and explore the problem from the value difference perspective via a novel insight on the value consistency across domains.

We introduce ODRL, the first benchmark tailored for evaluating off-dynamics RL methods where one needs to transfer policies across different domains with dynamics mismatch.

We propose a novel unsupervised skill discovery method through contrastive learning among behaviors, which makes the agent produce similar behaviors for the same skill and diverse behaviors for different skills.

We propose falSe COrrelation REduction (SCORE) for offline RL, a practically effective and theoretically provable algorithm.

We propose a novel single-stage privileged knowledge distillation method called the Historical Information Bottleneck (HIB) to narrow the sim-to-real gap for legged locomotion.

We propose Robust Offline Reinforcement Learning (RORL) with a novel conservative smoothing technique.

We study how RL can be empowered by contrastive learning in a class of Markov decision processes (MDPs) and Markov games (MGs) with low-rank transitions. For both models, we propose to extract the correct feature representations of the low-rank model by minimizing a contrastive loss.

We propose Pessimistic Bootstrapping for offline RL (PBRL), a purely uncertainty-driven offline algorithm without explicit policy constraints.

We propose a Dynamic Bottleneck (DB) model, which attains a dynamics-relevant representation based on the information-bottleneck principle.

We propose a principled exploration method for DRL through Optimistic Bootstrapping and Backward Induction (OB2I).