Selected Publications

Restoring missing gaps and intervals via global consistency and local coherence.

LabelDP Leaks Privacy - A Tightened Correlation-Aware Privacy Model for Labeled Training Data.

Differentially Private Zeroth-Order Methods for Scalable Large Language Model Fine-Tuning.

Stochastic Universal Adversarial Perturbations with Fixed Optimization Constraint and Ensured High-probability Transferability.

Lightweight Time Series Data Valuation on Time Series Foundation Models via In-Context Finetuning.

Safety at One Shot: Patching Fine-Tuned LLMs with A Single Instance.

Understanding and Preserving Safety in Fine-Tuned LLMs.

Time Series Reasoning via Process-Verifiable Thinking Data Synthesis and Scheduling for Tailored LLM Reasoning.

MsFormer: Enabling Robust Predictive Maintenance Services for Industrial Devices.

WaveMoE: A Wavelet-Enhanced Mixture-of-Experts Foundation Model for Time Series Forecasting.

Large language models for explainable fault diagnosis of machines.

Towards answering analytical query over hierarchical histogram under untrusted servers.

DuplexGuard: Safeguarding Deletion Right in Machine Unlearning via Duplex Watermarking.

Privacy Risks of Federated Knowledge Graph Embedding: New Membership Inference Attacks and Personalized Differential Privacy Defense.

FELEMN: Toward Efficient Feature-Level Machine Unlearning for Exact Privacy Protection.

Purifier$^{+}$: Plug-and-Play Backdoor Mitigation for Pre-Trained Models via Activation Alignment.

Zeroth-Order Federated Private Tuning for Pretrained Large Language Models.

Don't Say No: Jailbreaking LLM by Suppressing Refusal.

PreferCare: Preference Dataset Copyright Protection in LLM Alignment by Watermark Injection and Verification.

Integrating Time Series into LLMs via Multi-layer Steerable Embedding Fusion for Enhanced Forecasting.

Adversary-Aware DPO: Enhancing Safety Alignment in Vision Language Models via Adversarial Training.

Assessing Safety Risks and Quantization-aware Safety Patching for Quantized Large Language Models.

Clients Collaborate: Flexible Differentially Private Federated Learning with Guaranteed Improvement of Utility-Privacy Trade-off.

PoisonedEye: Knowledge Poisoning Attack on Retrieval-Augmented Generation based Large Vision-Language Models.

Contrastive Unlearning: A Contrastive Approach to Machine Unlearning.

Activation Approximations Can Incur Safety Vulnerabilities in Aligned LLMs: Comprehensive Analysis and Defense.

Activation Approximations Can Incur Safety Vulnerabilities Even in Aligned LLMs: Comprehensive Analysis and Defense.

SecPE: Secure Prompt Ensembling for Private and Robust Large Language Models.

SHAPE : Self-Improved Visual Preference Alignment by Iteratively Generating Holistic Winner.

Enhancing LLM Reasoning for Time Series Classification by Tailored Thinking and Fused Decision.

TSRating: Rating Quality of Diverse Time Series Data by Meta-learning from LLM Judgment.

Safeguarding Multimodal Knowledge Copyright in the RAG-as-a-Service Environment.

Q-resafe: Assessing Safety Risks and Quantization-aware Safety Patching for Quantized Large Language Models.

Module-Aware Parameter-Efficient Machine Unlearning on Transformers.

Towards Mitigating Excessive Forgetting in LLM Unlearning via Entanglement-Aware Unlearning with Proxy Constraint.

ParaAegis: Parallel Protection for Flexible Privacy-preserved Federated Learning.

Token-Level Inference-Time Alignment for Vision-Language Models.

Lightweight Time Series Data Valuation on Time Series Foundation Models via In-Context Finetuning.

Soft label enhanced graph neural network under heterophily.

Fostering Active Learning: A Study on Anonymous Q&A Software in Undergraduate Education.

LogExpertSolver: A Multi-Agent Framework for Domain-Specialized Log Parsing.

Cyber-Attack Detection and Localization for SCADA system of CPSs.

LicenseGPT: A Fine-tuned Foundation Model for Publicly Available Dataset License Compliance.

A Unified CatBoost Framework for IoT Data Classification Using Multi-Level Feature Extraction.

BACE-RUL: A Bi-directional Adversarial Network with Covariate Encoding for Machine Remaining Useful Life Prediction.

AnomalyGen: An Automated Semantic Log Sequence Generation Framework with LLM for Anomaly Detection.

ConfLogger: Enhance Systems' Configuration Diagnosability through Configuration Logging.

Generating High-Quality Datasets for Code Editing via Open-Source Language Models.

MaskArmor: Confidence masking-based defense mechanism for GNN against MIA.

Local Differentially Private Heavy Hitter Detection in Data Streams with Bounded Memory.

RemovalNet: DNN Fingerprint Removal Attacks.

Signed Graph Neural Ordinary Differential Equation for Modeling Continuous-Time Dynamics.

IGAMT: Privacy-Preserving Electronic Health Record Synthesization with Heterogeneity and Irregularity.

Cross-silo Federated Learning with Record-level Personalized Differential Privacy.

ERASER: Machine Unlearning in MLaaS via an Inference Serving-Aware Approach.

Physical Backdoor: Towards Temperature-based Backdoor Attacks in the Physical World.

Does Differential Privacy Prevent Backdoor Attacks in Practice?

SecPE: Secure Prompt Ensembling for Private and Robust Large Language Models.

PoisonPrompt: Backdoor Attack on Prompt-Based Large Language Models.

Balancing Generalization and Robustness in Adversarial Training via Steering through Clean and Adversarial Gradient Directions.

Rapid Plug-in Defenders.

PromptCARE: Prompt Copyright Protection by Watermark Injection and Verification.

DPAR: Decoupled Graph Neural Networks with Node-Level Differential Privacy.

Contrastive Unlearning: A Contrastive Approach to Machine Unlearning.

Clients Collaborate: Flexible Differentially Private Federated Learning with Guaranteed Improvement of Utility-Privacy Trade-off.

Differentially Private Zeroth-Order Methods for Scalable Large Language Model Finetuning.

eWAPA: An eBPF-based WASI Performance Analysis Framework for Web Assembly Runtimes.

GLA-DA: Global-Local Alignment Domain Adaptation for Multivariate Time Series.

Face It Yourselves: An LLM-Based Two-Stage Strategy to Localize Configuration Errors via Logs.

LLM-based Knowledge Pruning for Time Series Data Analytics on Edge-computing Devices.

eWAPA: An eBPF-based WASI Performance Analysis Framework for WebAssembly Runtimes.

FD-LLM: Large Language Model for Fault Diagnosis of Machines.

DP-starJ: A Differential Private Scheme towards Analytical Star-Join Queries.

Equitable Data Valuation Meets the Right to Be Forgotten in Model Markets.

FINER: Enhancing State-of-the-art Classifiers with Feature Attribution to Facilitate Security Analysis.

Closed-form Machine Unlearning for Matrix Factorization.

Personalized Differentially Private Federated Learning without Exposing Privacy Budgets.

Explaining Adversarial Robustness of Neural Networks from Clustering Effect Perspective.

MUter: Machine Unlearning on Adversarially Trained Models.

MULTIPAR: Supervised Irregular Tensor Factorization with Multi-task Learning for Computational Phenotyping.

ACQ: Few-shot Backdoor Defense via Activation Clipping and Quantizing.

Certified Minimax Unlearning with Generalization Rates and Deletion Capacity.

CAPP-130: A Corpus of Chinese Application Privacy Policy Summarization and Interpretation.

Interpretation Attacks and Defenses on Predictive Models Using Electronic Health Records.

Federated Semi-Supervised Learning with Annotation Heterogeneity.

Wasserstein Adversarial Examples on Univariant Time Series Data.

PromptCARE: Prompt Copyright Protection by Watermark Injection and Verification.

PoisonPrompt: Backdoor Attack on Prompt-based Large Language Models.

Does Differential Privacy Prevent Backdoor Attacks in Practice?

Local Differentially Private Heavy Hitter Detection in Data Streams with Bounded Memory.

ERASER: Machine Unlearning in MLaaS via an Inference Serving-Aware Approach.

Signed Graph Neural Ordinary Differential Equation for Modeling Continuous-time Dynamics.

Prompt Valuation Based on Shapley Values.

MAD-SGS: Multivariate Anomaly Detection with Multi-scale Self-learned Graph Structures.

CB-GAN: Generate Sensitive Data with a Convolutional Bidirectional Generative Adversarial Networks.

Fault Detection and Diagnosis of Air Handling Units Based On Adversarial Multi-task Learning.

Adversarial Maritime Trajectory Prediction with Real-time Spatial-Temporal Mutual Influence.

Robust Tensor SVD and Recovery With Rank Estimation.

Generating Adversarial Examples With Distance Constrained Adversarial Imitation Networks.

DP-HORUS: Differentially Private Hierarchical Count Histograms under Untrusted Server.

RobustFed: A Truth Inference Approach for Robust Federated Learning.

Higher-Order Masked Graph Neural Networks for Traffic Flow Prediction.

Purifier: Plug-and-play Backdoor Mitigation for Pre-trained Models Via Anomaly Activation Suppression.

Backdoor Attacks on Crowd Counting.

Vertical Federated Principal Component Analysis and Its Kernel Extension on Feature-wise Distributed Data.

MULTIPAR: Supervised Irregular Tensor Factorization with Multi-task Learning.

Towards Training Graph Neural Networks with Node-Level Differential Privacy.

Private Semi-supervised Knowledge Transfer for Deep Learning from Noisy Labels.

FedTracker: Furnishing Ownership Verification and Traceability for Federated Learning Model.

MAD-SGCN: Multivariate Anomaly Detection with Self-learning Graph Convolutional Networks.

Dealer: An End-to-End Model Marketplace with Differential Privacy.

Demonstration of Dealer: An End-to-End Model Marketplace with Differential Privacy.

Projected Federated Averaging with Heterogeneous Differential Privacy.

An Uplink Communication-Efficient Approach to Featurewise Distributed Sparse Optimization With Differential Privacy.

Temporal Network Embedding via Tensor Factorization.

Integer-arithmetic-only Certified Robustness for Quantized Neural Networks.

Communication Efficient Tensor Factorization for Decentralized Healthcare Networks.

Private Stochastic Non-convex Optimization with Improved Utility Rates.

Certified Robustness to Word Substitution Attack with Differential Privacy.

Vertical Federated Principal Component Analysis on Feature-Wise Distributed Data.

Communication Efficient Federated Generalized Tensor Factorization for Collaborative Health Data Analytics.

RobustFed: A Truth Inference Approach for Robust Federated Learning.

SemiFed: Semi-supervised Federated Learning with Consistency and Pseudo-Labeling.

SNEAK: Synonymous Sentences-Aware Adversarial Attack on Natural Language Video Localization.

Electrocardiogram Classification and Visual Diagnosis of Atrial Fibrillation with DenseECG.

Synergistic Generic Learning for Face Recognition From a Contaminated Single Sample per Person.

Robust Low-Rank Tensor Minimization via a New Tensor Spectral k-Support Norm.

Broadening Differential Privacy for Deep Learning Against Model Inversion Attacks.

Towards Training Robust Private Aggregation of Teacher Ensembles Under Noisy Labels.

Robust Irregular Tensor Factorization and Completion for Temporal Health Data Analysis.

Projection-free Online Empirical Risk Minimization with Privacy-preserving and Privacy Expiration.

Handling Incomplete Sensor Measurements in Fault Detection and Diagnosis for Building HVAC Systems.

VC-GAN: Classifying Vessel Types by Maritime Trajectories using Generative Adversarial Networks.