Yiwei Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yiwei Wang has a total of 32 co-authors. Here are the top co-authors by number of shared publications:
- Baolong Bi
Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China and University of Chinese Academy of Sciences, Beijing, China
2 shared publications Bryan Hooi
National University of Singapore, Singapore, Singapore
1 shared publication- Daniel Pérez Palomar
Hong Kong University of Science and Technology (HKUST), Kowloon, Hong Kong
1 shared publication - Hao Jiang
Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China
1 shared publication - Honglong Zhao
Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China
1 shared publication - Huaming Wu
Tianjin University, Tianjin, China
1 shared publication
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 5690-5699
STDR: Spatio-Temporal Decoupling for Real-Time Dynamic Scene Rendering
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 5690-5699
Open AccessSTDR: Spatio-Temporal Decoupling for Real-Time Dynamic Scene Rendering
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1262-1273
Towards Robust Heterogeneous Graph Explanations under Structural Perturbations
Explaining the decision-making process of Graph Neural Networks (GNNs) is essential for improving their transparency and reliability. However, real-world graphs are often heterogeneous and subject to structural noise, posing severe challenges to the robustness of existing explanation methods. To address these issues, we propose RoHeX, a Robust Heterogeneous GNN Explainer that enhances explanation quality under noisy conditions. RoHeX begins with a theoretical analysis revealing how different heterogeneous GNN architectures amplify structural perturbations through message passing. Building on this insight, we design a denoising variational inference framework that filters noisy structures and learns robust latent graph representations. Furthermore, we incorporate relation-aware heterogeneous semantics into the explanation generation process, formulating explanation as an optimization problem under the graph information bottleneck principle. This formulation enables RoHeX to balance fidelity and compactness, producing explanations that are both semantically meaningful and structurally stable. Comprehensive experiments on multiple real-world heterogeneous graphs demonstrate that RoHeX consistently surpasses state-of-the-art baselines in explanation fidelity, robustness to structural perturbations, and explainability.
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1262-1273
Open AccessTowards Robust Heterogeneous Graph Explanations under Structural Perturbations
Explaining the decision-making process of Graph Neural Networks (GNNs) is essential for improving their transparency and reliability. However, real-world graphs are often heterogeneous and subject to structural noise, posing severe challenges to the robustness of existing explanation methods. To address these issues, we propose RoHeX, a Robust Heterogeneous GNN Explainer that enhances explanation quality under noisy conditions. RoHeX begins with a theoretical analysis revealing how different heterogeneous GNN architectures amplify structural perturbations through message passing. Building on this insight, we design a denoising variational inference framework that filters noisy structures and learns robust latent graph representations. Furthermore, we incorporate relation-aware heterogeneous semantics into the explanation generation process, formulating explanation as an optimization problem under the graph information bottleneck principle. This formulation enables RoHeX to balance fidelity and compactness, producing explanations that are both semantically meaningful and structurally stable. Comprehensive experiments on multiple real-world heterogeneous graphs demonstrate that RoHeX consistently surpasses state-of-the-art baselines in explanation fidelity, robustness to structural perturbations, and explainability.
- Open Access
- 9 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Can Graph Descriptive Order Affect Solving Graph Problems with LLMs?
Large language models (LLMs) have achieved significant success in reasoning tasks, including mathematical reasoning and logical deduction. Among these reasoning tasks, graph problems stand out due to their complexity and unique structural characteristics, attracting considerable attention from researchers. Previous studies have explored LLMs' graph reasoning abilities through various techniques, such as different encoding methods for graph structures and the use of carefully designed prompts. However, a critical factor has been mostly overlooked: the prompt sequential order in which graph descriptions are presented to the models. In this study, we present the first comprehensive analysis of how the order of graph descriptions impacts LLM performance. Specifically, we comprehensively evaluate four graph description orders across six graph problems using six mainstream LLMs. The results reveal that: (1) ordered graph descriptions significantly improve LLMs' comprehension of graph structures; (2) the robustness of LLMs to graph description order varies across different tasks; and (3) the impact of graph order on performance is closely related to the inherent characteristics of tasks. This study provides a critical advancement in the application of LLMs for solving graph-related problems, paving the way for future research to optimize model performance through strategic graph description ordering.
- 9 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessCan Graph Descriptive Order Affect Solving Graph Problems with LLMs?
Large language models (LLMs) have achieved significant success in reasoning tasks, including mathematical reasoning and logical deduction. Among these reasoning tasks, graph problems stand out due to their complexity and unique structural characteristics, attracting considerable attention from researchers. Previous studies have explored LLMs' graph reasoning abilities through various techniques, such as different encoding methods for graph structures and the use of carefully designed prompts. However, a critical factor has been mostly overlooked: the prompt sequential order in which graph descriptions are presented to the models. In this study, we present the first comprehensive analysis of how the order of graph descriptions impacts LLM performance. Specifically, we comprehensively evaluate four graph description orders across six graph problems using six mainstream LLMs. The results reveal that: (1) ordered graph descriptions significantly improve LLMs' comprehension of graph structures; (2) the robustness of LLMs to graph description order varies across different tasks; and (3) the impact of graph order on performance is closely related to the inherent characteristics of tasks. This study provides a critical advancement in the application of LLMs for solving graph-related problems, paving the way for future research to optimize model performance through strategic graph description ordering.
- Open Access
- 14 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Vulnerability of Large Language Models to Output Prefix Jailbreaks: Impact of Positions on Safety
Previous research on jailbreak attacks has mainly focused on optimizing the adversarial snippet content injected into input prompts to expose LLM security vulnerabilities. A significant portion of this research focuses on developing more complex, less readable adversarial snippets that can achieve higher attack success rates. In contrast to this trend, our research investigates the impact of the adversarial snippet's position on the effectiveness of jailbreak attacks. We find that placing a simple and readable adversarial snippet at the beginning of the output effectively exposes LLM safety vulnerabilities, leading to much higher attack success rates than the input suffix attack or promptbased output jailbreaks. Precisely speaking, we discover that directly enforcing the user's target embedded output prefix is an effective method to expose LLMs' safety vulnerabilities.
- 14 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessVulnerability of Large Language Models to Output Prefix Jailbreaks: Impact of Positions on Safety
Previous research on jailbreak attacks has mainly focused on optimizing the adversarial snippet content injected into input prompts to expose LLM security vulnerabilities. A significant portion of this research focuses on developing more complex, less readable adversarial snippets that can achieve higher attack success rates. In contrast to this trend, our research investigates the impact of the adversarial snippet's position on the effectiveness of jailbreak attacks. We find that placing a simple and readable adversarial snippet at the beginning of the output effectively exposes LLM safety vulnerabilities, leading to much higher attack success rates than the input suffix attack or promptbased output jailbreaks. Precisely speaking, we discover that directly enforcing the user's target embedded output prefix is an effective method to expose LLMs' safety vulnerabilities.
- Open Access
- 27 February 2023
- Published by Association for Computing Machinery (ACM)
- p. 348-356
Graph Explicit Neural Networks: Explicitly Encoding Graphs for Efficient and Accurate Inference
As the state-of-the-art graph learning models, the message passing based neural networks (MPNNs) implicitly use the graph topology as the "pathways" to propagate node features. This implicit use of graph topology induces the MPNNs' over-reliance on (node) features and high inference latency, which hinders their large-scale applications in industrial contexts. To mitigate these weaknesses, we propose the Graph Explicit Neural Network (GENN) framework. GENN can be flexibly applied to various MPNNs and improves them by providing more efficient and accurate inference that is robust in feature-constrained settings. Specifically, we carefully incorporate recent developments in network embedding methods to efficiently prioritize the graph topology for inference. From this vantage, GENN explicitly encodes the topology as an important source of information to mitigate the reliance on node features. Moreover, by adopting knowledge distillation (KD) techniques, GENN takes an MPNN as the teacher to supervise the training for better effectiveness while avoiding the teacher's high inference latency. Empirical results show that our GENN infers dramatically faster than its MPNN teacher by 40x-78x. In terms of accuracy, GENN yields significant gains (more than 40%) for its MPNN teacher when the node features are limited based on our explicit encoding. Moreover, GENN outperforms the MPNN teacher even in feature-rich settings thanks to our KD design.
- 27 February 2023
- Published by Association for Computing Machinery (ACM)
- p. 348-356
Open AccessGraph Explicit Neural Networks: Explicitly Encoding Graphs for Efficient and Accurate Inference
As the state-of-the-art graph learning models, the message passing based neural networks (MPNNs) implicitly use the graph topology as the "pathways" to propagate node features. This implicit use of graph topology induces the MPNNs' over-reliance on (node) features and high inference latency, which hinders their large-scale applications in industrial contexts. To mitigate these weaknesses, we propose the Graph Explicit Neural Network (GENN) framework. GENN can be flexibly applied to various MPNNs and improves them by providing more efficient and accurate inference that is robust in feature-constrained settings. Specifically, we carefully incorporate recent developments in network embedding methods to efficiently prioritize the graph topology for inference. From this vantage, GENN explicitly encodes the topology as an important source of information to mitigate the reliance on node features. Moreover, by adopting knowledge distillation (KD) techniques, GENN takes an MPNN as the teacher to supervise the training for better effectiveness while avoiding the teacher's high inference latency. Empirical results show that our GENN infers dramatically faster than its MPNN teacher by 40x-78x. In terms of accuracy, GENN yields significant gains (more than 40%) for its MPNN teacher when the node features are limited based on our explicit encoding. Moreover, GENN outperforms the MPNN teacher even in feature-rich settings thanks to our KD design.
- 27 June 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 67 (16) , 4231-4244
Optimization Algorithms for Graph Laplacian Estimation via ADMM and MM
In this paper, we study the graph Laplacian estimation problem under a given connectivity topology. We aim at enriching the unified graph learning framework proposed by Egilmez et al. and improve the optimality performance of the Combinatorial Graph Laplacian (CGL) case. We apply the well-known Alternating Direction Method of Multipliers (ADMM) and Majorization-Minimization (MM) algorithmic frameworks and propose two algorithms, namely, GLE-ADMM and GLE-MM, for graph Laplacian estimation. Both algorithms can achieve an optimality gap as low as $10^{-4}$ , around three orders of magnitude more accurate than the benchmark. In addition, we find that GLE-ADMM is more computationally efficient in a dense topology (e.g., an almost complete graph) while GLE-MM is more suitable for sparse graphs (e.g., trees). Furthermore, we consider exploiting the leading eigenvectors of the sample covariance matrix as a nominal eigensubspace and propose a third algorithm named GLENE, which is also based on ADMM. Numerical experiments show that the inclusion of a nominal eigensubspace significantly improves the estimation of the graph Laplacian when the sample size is smaller than or comparable to the problem dimension.- 27 June 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 67 (16) , 4231-4244
Optimization Algorithms for Graph Laplacian Estimation via ADMM and MM
In this paper, we study the graph Laplacian estimation problem under a given connectivity topology. We aim at enriching the unified graph learning framework proposed by Egilmez et al. and improve the optimality performance of the Combinatorial Graph Laplacian (CGL) case. We apply the well-known Alternating Direction Method of Multipliers (ADMM) and Majorization-Minimization (MM) algorithmic frameworks and propose two algorithms, namely, GLE-ADMM and GLE-MM, for graph Laplacian estimation. Both algorithms can achieve an optimality gap as low as $10^{-4}$ , around three orders of magnitude more accurate than the benchmark. In addition, we find that GLE-ADMM is more computationally efficient in a dense topology (e.g., an almost complete graph) while GLE-MM is more suitable for sparse graphs (e.g., trees). Furthermore, we consider exploiting the leading eigenvectors of the sample covariance matrix as a nominal eigensubspace and propose a third algorithm named GLENE, which is also based on ADMM. Numerical experiments show that the inclusion of a nominal eigensubspace significantly improves the estimation of the graph Laplacian when the sample size is smaller than or comparable to the problem dimension.