45 citations · 113 across the 28 of their papers we have counts for
28 papers
Character as a Latent Variable in Large Language Models: A Mechanistic Account of Emergent Misalignment and Conditional Safety Failures
Yanghao Su, Wenbo Zhou, Tianwei Zhang +4
Emergent Misalignment refers to a failure mode in which fine-tuning large language models (LLMs) on narrowly scoped data induces broadly misaligned behavior. Prior explanations mai…
EditMark: Watermarking Large Language Models based on Model Editing
Shuai Li, Kejiang Chen, Jun Jiang +5
Large Language Models (LLMs) have demonstrated remarkable capabilities, but their training requires extensive data and computational resources, rendering them valuable digital asse…
SafeGuider: Robust and Practical Content Safety Control for Text-to-Image Models
Peigui Qi, Kunsheng Tang, Wenbo Zhou +5
Text-to-image models have shown remarkable capabilities in generating high-quality images from natural language descriptions. However, these models are highly vulnerable to adversa…
Clean Image May be Dangerous: Data Poisoning Attacks Against Deep Hashing
Shuai Li, Jie Zhang, Yuang Qi +4
Large-scale image retrieval using deep hashing has become increasingly popular due to the exponential growth of image data and the remarkable feature extraction capabilities of dee…
Exploiting Vulnerabilities in Speech Translation Systems through Targeted Adversarial Attacks
Chang Liu, Haolin Wu, Xi Yang +7
As speech translation (ST) systems become increasingly prevalent, understanding their vulnerabilities is crucial for ensuring robust and reliable communication. However, limited wo…
EraseAnything: Enabling Concept Erasure in Rectified Flow Transformers
Daiheng Gao, Shilin Lu, Shaw Walters +8
Removing unwanted concepts from large-scale text-to-image (T2I) diffusion models while maintaining their overall generative quality remains an open challenge. This difficulty is es…
On the Vulnerability of Text Sanitization
Meng Tong, Kejiang Chen, Xiaojian Yuan +4
Text sanitization, which employs differential privacy to replace sensitive tokens with new ones, represents a significant technique for privacy protection. Typically, its performan…
GenderCARE: A Comprehensive Framework for Assessing and Reducing Gender Bias in Large Language Models
Kunsheng Tang, Wenbo Zhou, Jie Zhang +7
Large language models (LLMs) have exhibited remarkable capabilities in natural language generation, but they have also been observed to magnify societal biases, particularly those…
AquaLoRA: Toward White-box Protection for Customized Stable Diffusion Models via Watermark LoRA
Weitao Feng, Wenbo Zhou, Jiyan He +6
Diffusion models have achieved remarkable success in generating high-quality images. Recently, the open-source models represented by Stable Diffusion (SD) are thriving and are acce…
Model X-ray:Detecting Backdoored Models via Decision Boundary
Yanghao Su, Jie Zhang, Ting Xu +3
Backdoor attacks pose a significant security vulnerability for deep neural networks (DNNs), enabling them to operate normally on clean inputs but manipulate predictions when specif…
Data-Free Hard-Label Robustness Stealing Attack
Xiaojian Yuan, Kejiang Chen, Wen Huang +3
The popularity of Machine Learning as a Service (MLaaS) has led to increased concerns about Model Stealing Attacks (MSA), which aim to craft a clone model by querying MLaaS. Curren…
Controlling risks of AI in chemical science with agents
Jiyan He, Weitao Feng, Haoxiang Guan +11
Artificial intelligence is rapidly advancing scientific discovery, but this progress carries risks of misuse, such as the creation of harmful substances, or circumvention of establ…
Detecting Voice Cloning Attacks via Timbre Watermarking
Chang Liu, Jie Zhang, Tianwei Zhang +3
Nowadays, it is common to release audio content to the public. However, with the rise of voice cloning technology, attackers have the potential to easily impersonate a specific per…
Improving Adversarial Transferability by Stable Diffusion
Jiayang Liu, Siyu Zhu, Siyuan Liang +4
Deep neural networks (DNNs) are susceptible to adversarial examples, which introduce imperceptible perturbations to benign samples, deceiving DNN predictions. While some attack met…
Turning Your Strength into Watermark: Watermarking Large Language Model via Knowledge Injection
Shuai Li, Kejiang Chen, Kunsheng Tang +4
Large language models (LLMs) have demonstrated outstanding performance, making them valuable digital assets with significant commercial potential. Unfortunately, the LLM and its AP…
Segue: Side-information Guided Generative Unlearnable Examples for Facial Privacy Protection in Real World
Zhiling Zhang, Jie Zhang, Kui Zhang +3
The widespread use of face recognition technology has given rise to privacy concerns, as many individuals are worried about the collection and utilization of their facial data. To…
InferDPT: Privacy-Preserving Inference for Closed-box Large Language Model
Meng Tong, Kejiang Chen, Jie Zhang +5
Large language models (LLMs), like ChatGPT, have greatly simplified text generation tasks. However, they have also raised concerns about privacy risks such as data leakage and unau…
Catch You Everything Everywhere: Guarding Textual Inversion via Concept Watermarking
Weitao Feng, Jiyan He, Jie Zhang +4
AIGC (AI-Generated Content) has achieved tremendous success in many applications such as text-to-image tasks, where the model can generate high-quality images with diverse prompts,…
Aparecium: Revealing Secrets from Physical Photographs
Zhe Lei, Jie Zhang, Jingtao Li +2
Watermarking is a crucial tool for safeguarding copyrights and can serve as a more aesthetically pleasing alternative to QR codes. In recent years, watermarking methods based on de…
Watermarking Text Generated by Black-Box Language Models
Xi Yang, Kejiang Chen, Weiming Zhang +5
LLMs now exhibit human-like skills in various fields, leading to worries about misuse. Thus, detecting generated text is crucial. However, passive detection methods are stuck in do…
Pseudo Label-Guided Model Inversion Attack via Conditional Generative Adversarial Network
Xiaojian Yuan, Kejiang Chen, Jie Zhang +3
Model inversion (MI) attacks have raised increasing concerns about privacy, which can reconstruct training data from public models. Indeed, MI attacks can be formalized as an optim…
DeAR: A Deep-learning-based Audio Re-recording Resilient Watermarking
Chang Liu, Jie Zhang, Han Fang +3
Audio watermarking is widely used for leaking source tracing. The robustness of the watermark determines the traceability of the algorithm. With the development of digital technolo…
Ada3Diff: Defending against 3D Adversarial Point Clouds via Adaptive Diffusion
Kui Zhang, Hang Zhou, Jie Zhang +3
Deep 3D point cloud models are sensitive to adversarial attacks, which poses threats to safety-critical applications such as autonomous driving. Robust training and defend-by-denoi…
Tracing Text Provenance via Context-Aware Lexical Substitution
Xi Yang, Jie Zhang, Kejiang Chen +4
Text content created by humans or language models is often stolen or misused by adversaries. Tracing text provenance can help claim the ownership of text content or identify the ma…