7 citations · 15 across the 18 of their papers we have counts for
20 papers
GradSentry: Gradient Spectral Entropy for Backdoor Sample Filtering in Large Language Model Fine-Tuning
Haodong Zhao, Tianyi Xu, Tianhang Zhao +2
Fine-tuning Large Language Models with untrusted data exposes models to backdoor attacks, where poisoned samples cause targeted misbehavior. Existing sample-filtering defenses rely…
EmbTracker: Traceable Black-box Watermarking for Federated Language Models
Haodong Zhao, Jinming Hu, Yijie Bai +6
Federated Language Model (FedLM) allows a collaborative learning without sharing raw data, yet it introduces a critical vulnerability, as every untrustworthy client may leak the re…
ProtegoFed: Backdoor-Free Federated Instruction Tuning with Interspersed Poisoned Data
Haodong Zhao, Jinming Hu, Zhaomin Wu +7
Federated Instruction Tuning (FIT) enables collaborative instruction tuning of large language models across multiple organizations (clients) in a cross-silo setting without requiri…
Revisiting Backdoor Threat in Federated Instruction Tuning from a Signal Aggregation Perspective
Haodong Zhao, Jinming Hu, Gongshen Liu
Federated learning security research has predominantly focused on backdoor threats from a minority of malicious clients that intentionally corrupt model updates. This paper challen…
Patronus: Identifying and Mitigating Transferable Backdoors in Pre-trained Language Models
Tianhang Zhao, Haodong Zhao, Wei Du +5
The ``Pre-train, then fine-tune'' paradigm has revolutionized Natural Language Processing (NLP). In this context, transferable backdoors pose a severe threat to the Pre-trained Lan…
Judge Before Answer: Can MLLM Discern the False Premise in Question?
Jidong Li, Lingyong Fang, Haodong Zhao +2
Multimodal large language models (MLLMs) have witnessed astonishing advancements in recent years. Despite these successes, MLLMs remain vulnerable to flase premise problems. Howeve…
GhostEI-Bench: Do Mobile Agents Resilience to Environmental Injection in Dynamic On-Device Environments?
Chiyu Chen, Xinhao Song, Yunkai Chai +7
Vision-Language Models (VLMs) are increasingly deployed as autonomous agents to navigate mobile graphical user interfaces (GUIs). Operating in dynamic on-device ecosystems, which i…
Thinking in a Crowd: How Auxiliary Information Shapes LLM Reasoning
Haodong Zhao, Chenyan Zhao, Yansi Li +2
The capacity of Large Language Models (LLMs) to reason is fundamental to their application in complex, knowledge-intensive domains. In real-world scenarios, LLMs are often augmente…
LLM DNA: Tracing Model Evolution via Functional Representations
Zhaomin Wu, Haodong Zhao, Ziyang Wang +3
The explosive growth of large language models (LLMs) has created a vast but opaque landscape: millions of models exist, yet their evolutionary relationships through fine-tuning, di…
Reasoning or Rambling? Exploring the Effect of Thinking on Agent Persuasion
Haodong Zhao, Jidong Li, Zhaomin Wu +4
Understanding persuasion is critical for the safety and reliability of multi-agent systems built on large language models (LLMs). This paper studies persuasion dynamics by contrast…
FedRS-Bench: Realistic Federated Learning Datasets and Benchmarks in Remote Sensing
Haodong Zhao, Peng Peng, Chiyu Chen +2
Remote sensing (RS) images are usually produced at an unprecedented scale, yet they are geographically and institutionally distributed, making centralized model training challengin…
Watch Out Your Album! On the Inadvertent Privacy Memorization in Multi-Modal Large Language Models
Tianjie Ju, Yi Hua, Hao Fei +7
Multi-Modal Large Language Models (MLLMs) have exhibited remarkable performance on various vision-language tasks such as Visual Question Answering (VQA). Despite accumulating evide…
Vertical Federated Learning in Practice: The Good, the Bad, and the Ugly
Zhaomin Wu, Zhen Qin, Junyi Hou +4
Vertical Federated Learning (VFL) is a privacy-preserving collaborative learning paradigm that enables multiple parties with distinct feature sets to jointly train machine learning…
NSmark: Null Space Based Black-box Watermarking Defense Framework for Language Models
Haodong Zhao, Jinming Hu, Peixuan Li +6
Language models (LMs) have emerged as critical intellectual property (IP) assets that necessitate protection. Although various watermarking strategies have been proposed, they rema…
Flooding Spread of Manipulated Knowledge in LLM-Based Multi-Agent Communities
Tianjie Ju, Yiting Wang, Xinbei Ma +7
The rapid adoption of large language models (LLMs) in multi-agent systems has highlighted their impressive capabilities in various applications, such as collaborative problem-solvi…
Revisiting the Information Capacity of Neural Network Watermarks: Upper Bound Estimation and Beyond
Fangqi Li, Haodong Zhao, Wei Du +1
To trace the copyright of deep neural networks, an owner can embed its identity information into its model as a watermark. The capacity of the watermark quantify the maximal volume…
Backdoor Attacks and Countermeasures in Natural Language Processing Models: A Comprehensive Security Review
Pengzhou Cheng, Zongru Wu, Wei Du +3
Language Models (LMs) are becoming increasingly popular in real-world applications. Outsourcing model training and data hosting to third-party platforms has become a standard metho…
UOR: Universal Backdoor Attacks on Pre-trained Language Models
Wei Du, Peixuan Li, Boqun Li +2
Backdoors implanted in pre-trained language models (PLMs) can be transferred to various downstream tasks, which exposes a severe security threat. However, most existing backdoor at…
A Universal Identity Backdoor Attack against Speaker Verification based on Siamese Network
Haodong Zhao, Wei Du, Junjie Guo +1
Speaker verification has been widely used in many authentication scenarios. However, training models for speaker verification requires large amounts of data and computing power, so…
FedPrompt: Communication-Efficient and Privacy Preserving Prompt Tuning in Federated Learning
Haodong Zhao, Wei Du, Fangqi Li +2
Federated learning (FL) has enabled global model training on decentralized data in a privacy-preserving way by aggregating model updates. However, for many natural language process…