3

MASQUE: Localized Adversarial Makeup Generation with Text-Guided Diffusion Models

We introduce MASQUE, a diffusion-based framework that generates localized adversarial makeup guided by user-defined text prompts.

Youngjin Kwon, Xiao Zhang

Generalizable Targeted Data Poisoning against Varying Physical Objects

We take the first step toward understanding the real-world threats of TDP by studying its generalizability across varying physical conditions.

Zhizhen Chen, Zhengyu Zhao, Subrat Kishore Dutta, Chenhao Lin, Chao Shen, Xiao Zhang

DiffCAP: Diffusion-based Cumulative Adversarial Purification for Vision Language Models

This paper introduces DiffCAP, a novel diffusion-based purification strategy that can effectively neutralize adversarial corruptions in VLMs.

Jia Fu, Yongtao Wu, Yihang Chen, Kunyu Peng, Xiao Zhang, Volkan Cevher, Sepideh Pashami, Anders Holst

GASP: Efficient Black-Box Generation of Adversarial Suffixes for Jailbreaking LLMs

We introduce Generative Adversarial Suffix Prompter (GASP), a novel framework that combines human-readable prompt generation with Latent Bayesian Optimization (LBO) to improve adversarial suffix creation in a fully black-box setting.

Advik Raj Basani, Xiao Zhang

Predicting Time-varying Flux and Balance in Metabolic Systems using Structured Neural ODE Processes

We propose a structured neural ODE process model to estimate flux and balance samples using gene-expression time-series data

Santanu Rathod, Pietro Liò, Xiao Zhang

Invisibility Cloak: Disappearance under Human Pose Estimation via Backdoor Attacks

We propose a general framework to craft invisibility cloak for human pose estimation models

Minxing Zhang, Michael Backes, Xiao Zhang

Improving the Efficiency of Self-Supervised Adversarial Training through Latent Clustering-based Selection

We introduce a Latent Clustering-based Selection method to choose a core subset from the entire unlabeled dataset, aiming to improve the efficiency of self-supervised adversarial training while preserving robustness.

Somrita Ghosh, Yuelin Xu, Xiao Zhang

Understanding Adversarially Robust Generalization via Weight-Curvature Index

We introduce the Weight-Curvature Index (WCI), a novel metric that captures the interplay between model parameters and loss landscape curvature to better understand and improve adversarially robust generalization in deep learning.

Yuelin Xu, Xiao Zhang

AutoDefense: Multi-Agent LLM Defense against Jailbreak Attacks

We propose AutoDefense, a response-filtering based multi-agent defense framework that filters harmful responses from LLMs.

Yifan Zeng, Yiran Wu, Xiao Zhang, Huazheng Wang, Qingyun Wu

Transferable Availability Poisoning Attacks

We propose an availability poisoning attack for generating transferable poisoned data across different victim learners.

Yiyong Liu, Michael Backes, Xiao Zhang