Adversarial Learning

Multi-Turn Jailbreaking Large Language Models via Attention Shifting AAAI 2025 NEXUS: Network Exploration for eXploiting Unsafe Sequences in Multi-Turn LLM Jailbreaks EMNLP 2025 Novel Loss-Enhanced Universal Adversarial Patches for Sustainable Speaker Privacy Interspeech 2025 Nullspace Disentanglement for Red Teaming Language Models EMNLP 2025 NullSwap: Proactive Identity Cloaking Against Deepfake Face Swapping ICCV 2025 NumbOD: A Spatial-Frequency Fusion Attack Against Object Detectors AAAI 2025 ODDR: Outlier Detection & Dimension Reduction Based Defense Against Adversarial Patches ICCV 2025 On Minimizing Adversarial Counterfactual Error in Adversarial Reinforcement Learning ICLR 2025 On the Relationship Between Robustness and Expressivity of Graph Neural Networks AISTATS 2025 On the Risk of Evidence Pollution for Malicious Social Text Detection in the Era of LLMs ACL 2025 On the Robustness of Distributed Machine Learning Against Transfer Attacks AAAI 2025 On the Zero-shot Adversarial Robustness of Vision-Language Models: A Truly Zero-shot and Training-free Approach CVPR 2025 PBCAT: Patch-Based Composite Adversarial Training against Physically Realizable Attacks on Object Detection ICCV 2025 Perception-Guided Jailbreak Against Text-to-Image Models AAAI 2025 Phi: Preference Hijacking in Multi-modal Large Language Models at Inference Time EMNLP 2025 PhyCamo: A Robust Physical Camouflage via Contrastive Learning for Multi-View Physical Adversarial Attack AAAI 2025 PIGuard: Prompt Injection Guardrail via Mitigating Overdefense for Free ACL 2025 Pixel Is Not a Barrier: An Effective Evasion Attack for Pixel-Domain Diffusion Models AAAI 2025 PLA: Prompt Learning Attack against Text-to-Image Generative Models ICCV 2025 Playing the Fool: Jailbreaking LLMs and Multimodal LLMs with Out-of-Distribution Strategy CVPR 2025 Poisoning Bayesian Inference via Data Deletion and Replication AISTATS 2025 PoolAtnRes: Towards Generalisable Differential Morphing Attack Detection WACV 2025 Power of Diversity: Enhancing Data-Free Black-Box Attack with Domain-Augmented Learning AAAI 2025 Pretend Benign: A Stealthy Adversarial Attack by Exploiting Vulnerabilities in Cooperative Perception ICCV 2025 Pre-Trained Multiple Latent Variable Generative Models are Good Defenders Against Adversarial Attacks WACV 2025