8 citations · 14 across the 9 of their papers we have counts for
10 papers
Probe-based Fine-tuning for Reducing Toxicity
Jan Wehner, Mario Fritz
Probes trained on model activations can detect undesirable behaviors like deception or biases that are difficult to identify from outputs alone. This makes them useful detectors to…
DP-SNP-TIHMM: Differentially Private, Time-Inhomogeneous Hidden Markov Models for Synthesizing Genome-Wide Association Datasets
Shadi Rahimian, Mario Fritz
Single nucleotide polymorphism (SNP) datasets are fundamental to genetic studies but pose significant privacy risks when shared. The correlation of SNPs with each other makes stron…
Deepfake Detection that Generalizes Across Benchmarks
Andrii Yermakov, Jan Cech, Jiri Matas +1
The generalization of deepfake detectors to unseen manipulation techniques remains a challenge for practical deployment. Although many approaches adapt foundation models by introdu…
Stealix: Model Stealing via Prompt Evolution
Zhixiong Zhuang, Hui-Po Wang, Maria-Irina Nicolae +1
Model stealing poses a significant security risk in machine learning by enabling attackers to replicate a black-box model without access to its training data, thus jeopardizing int…
ProxyPrompt: Securing System Prompts against Prompt Extraction Attacks
Zhixiong Zhuang, Maria-Irina Nicolae, Hui-Po Wang +1
The integration of large language models (LLMs) into a wide range of applications has highlighted the critical role of well-crafted system prompts, which require extensive testing…
Medical Multimodal Model Stealing Attacks via Adversarial Domain Alignment
Yaling Shen, Zhixiong Zhuang, Kun Yuan +4
Medical multimodal large language models (MLLMs) are becoming an instrumental part of healthcare systems, assisting medical personnel with decision making and results analysis. Mod…
Fundamental Risks in the Current Deployment of General-Purpose AI Models: What Have We (Not) Learnt From Cybersecurity?
Mario Fritz
General Purpose AI - such as Large Language Models (LLMs) - have seen rapid deployment in a wide range of use cases. Most surprisingly, they have have made their way from plain lan…
Red-Teaming Text-to-Image Models via In-Context Experience Replay and Semantic-Preserving Prompt Rewriting
Zhi-Yi Chin, Pin-Yu Chen, Wei-Chen Chiu +1
Understanding the capabilities of text-to-image (T2I) models in harmful content generation is essential to safety and compliance. However, human red-teaming is costly and inconsist…
LLM2Swarm: Robot Swarms that Responsively Reason, Plan, and Collaborate through LLMs
Volker Strobel, Marco Dorigo, Mario Fritz
Robot swarms are composed of many simple robots that communicate and collaborate to fulfill complex tasks. Robot controllers usually need to be specified by experts on a case-by-ca…
Stealthy Imitation: Reward-guided Environment-free Policy Stealing
Zhixiong Zhuang, Maria-Irina Nicolae, Mario Fritz
Deep reinforcement learning policies, which are integral to modern control systems, represent valuable intellectual property. The development of these policies demands considerable…