Search Results for author:
Found 7 papers, 3 papers with code
Stable Signature is Unstable: Removing Image Watermark from Diffusion Models
Our results show that our attack can effectively remove the watermark from a diffusion model such that its generated images are non-watermarked, while maintaining the visual quality of the generated images.
Concealing Backdoor Model Updates in Federated Learning by Trigger-Optimized Data Poisoning
Federated Learning (FL) is a decentralized machine learning method that enables participants to collaboratively train a model without sharing their private data.
PLeak: Prompt Leaking Attacks against Large Language Model Applications
As a result, a natural attack, called prompt leaking, is to steal the system prompt from an LLM application, which compromises the developer's intellectual property.
A Transfer Attack to Image Watermarks
The robustness of such watermark-based detector against evasion attacks in the white-box and black-box settings is well understood in the literature.
GradSafe: Detecting Unsafe Prompts for LLMs via Safety-Critical Gradient Analysis
In this study, we propose GradSafe, which effectively detects unsafe prompts by scrutinizing the gradients of safety-critical parameters in LLMs.
Mendata: A Framework to Purify Manipulated Training Data
Untrusted data used to train a model might have been manipulated to endow the learned model with hidden properties that the data contributor might later exploit.
SneakyPrompt: Jailbreaking Text-to-image Generative Models
Text-to-image generative models such as Stable Diffusion and DALL$\cdot$E raise many ethical concerns due to the generation of harmful images such as Not-Safe-for-Work (NSFW) ones.
