arXiv subjects

Yixiang Shan

Publications and source records attributed to Yixiang Shan.

5 recordsLinked to original sources

arXiv2026

Decoupling Policy Extraction for Offline Reinforcement Learning

Offline RL methods commonly jointly train the actor and critic, where the critic is used to guide the actor toward higher-value actions. This coupled learning process is well motivated in online RL, where an improved actor collects new data that can further update the actor and the critic. However, training data remains fixed in offline RL, making actor-side policy improvement unable to generate new data to validate or correct the critic. Moreover, retaining this coupled paradigm leads to two related challenges. Firstly, actor updates can drift toward high-valued but potentially out-of-distribution (OOD) actions and amplify critic overestimation. Secondly, conservative value estimation or behavior-cloning regularization creates a difficult trade-off between suppressing OOD actions and selecting high-value actions within the data-supported region. Motivated by this observation, we revisit the conventional offline RL paradigm and propose decoupling policy improvement from actor training. Specifically, we train the actor solely to model the behavior distribution and perform policy improvement at inference time by reranking multiple actor-generated proposals with a separately learned critic. We refer to this paradigm as the decoupled policy extraction paradigm. Under such paradigm, the actor provides behavior-supported action candidates, while the critic performs value-based selection within this candidate set. Extensive experiments show that the decoupled policy extraction paradigm outperforms both behavior cloning and jointly learned offline RL methods, while remaining effective even with a naive Q-learning critic.

cs.LG↗

arXiv2025

RAD: Retrieval High-quality Demonstrations to Enhance Decision-making

Offline reinforcement learning (RL) learns policies from fixed datasets, thereby avoiding costly or unsafe environment interactions. However, its reliance on finite static datasets inherently restricts the ability to generalize beyond the training distribution. Prior solutions based on synthetic data augmentation often fail to generalize to unseen scenarios in the (augmented) dataset. To address these challenges, we propose Retrieval High-quAlity Demonstrations (RAD) for decision-making, which innovatively introduces a retrieval mechanism into offline RL. Specifically, RAD retrieves high-return and reachable states from the offline dataset as target states, and leverages a generative model to generate sub-trajectories conditioned on these targets for planning. Since the targets are high-return states, once the agent reaches such a target, it can continue to obtain high returns by following the associated high-return actions, thereby improving policy generalization. Extensive experiments confirm that RAD achieves competitive or superior performance compared to baselines across diverse benchmarks, validating its effectiveness. Our code is available at https://github.com/LeahGL/RAD.

cs.AI↗

arXiv2024

Contrastive Diffuser: Planning Towards High Return States via Contrastive Learning

The performance of offline reinforcement learning (RL) is sensitive to the proportion of high-return trajectories in the offline dataset. However, in many simulation environments and real-world scenarios, there are large ratios of low-return trajectories rather than high-return trajectories, which makes learning an efficient policy challenging. In this paper, we propose a method called Contrastive Diffuser (CDiffuser) to make full use of low-return trajectories and improve the performance of offline RL algorithms. Specifically, CDiffuser groups the states of trajectories in the offline dataset into high-return states and low-return states and treats them as positive and negative samples correspondingly. Then, it designs a contrastive mechanism to pull the trajectory of an agent toward high-return states and push them away from low-return states. Through the contrast mechanism, trajectories with low returns can serve as negative examples for policy learning, guiding the agent to avoid areas associated with low returns and achieve better performance. Experiments on 14 commonly used D4RL benchmarks demonstrate the effectiveness of our proposed method. Our code is publicly available at \url{https://anonymous.4open.science/r/CDiffuser}.

cs.LG↗

arXiv2024

DiffStitch: Boosting Offline Reinforcement Learning with Diffusion-based Trajectory Stitching

In offline reinforcement learning (RL), the performance of the learned policy highly depends on the quality of offline datasets. However, in many cases, the offline dataset contains very limited optimal trajectories, which poses a challenge for offline RL algorithms as agents must acquire the ability to transit to high-reward regions. To address this issue, we introduce Diffusion-based Trajectory Stitching (DiffStitch), a novel diffusion-based data augmentation pipeline that systematically generates stitching transitions between trajectories. DiffStitch effectively connects low-reward trajectories with high-reward trajectories, forming globally optimal trajectories to address the challenges faced by offline RL algorithms. Empirical experiments conducted on D4RL datasets demonstrate the effectiveness of DiffStitch across RL methodologies. Notably, DiffStitch demonstrates substantial enhancements in the performance of one-step methods (IQL), imitation learning methods (TD3+BC), and trajectory optimization methods (DT).

cs.LG↗

arXiv2022

Learning the Network of Graphs for Graph Neural Networks

Graph neural networks (GNNs) have achieved great success in many scenarios with graph-structured data. However, in many real applications, there are three issues when applying GNNs: graphs are unknown, nodes have noisy features, and graphs contain noisy connections. Aiming at solving these problems, we propose a new graph neural network named as GL-GNN. Our model includes multiple sub-modules, each sub-module selects important data features and learn the corresponding key relation graph of data samples when graphs are unknown. GL-GNN further obtains the network of graphs by learning the network of sub-modules. The learned graphs are further fused using an aggregation method over the network of graphs. Our model solves the first issue by simultaneously learning multiple relation graphs of data samples as well as a relation network of graphs, and solves the second and the third issue by selecting important data features as well as important data sample relations. We compare our method with 14 baseline methods on seven datasets when the graph is unknown and 11 baseline methods on two datasets when the graph is known. The results show that our method achieves better accuracies than the baseline methods and is capable of selecting important features and graph edges from the dataset. Our code will be publicly available at \url{https://github.com/Looomo/GL-GNN}.

cs.LG↗