6 citations · 7 across the 10 of their papers we have counts for
12 papers
Towards Safe Autonomy for Computer Use
Dominik Seip, Bartol Markovinović, Matthias Hein
The safety of LLM-based agents currently rests on alignment training and safety classifiers, neither of which can rule out that a sophisticated indirect prompt injection (IPI) goes…
Preference Redirection via Attention Concentration: An Attack on Computer Use Agents
Dominik Seip, Matthias Hein
Advancements in multimodal foundation models have enabled the development of Computer Use Agents (CUAs) capable of autonomously interacting with GUI environments. As CUAs are not r…
Visual Memory Injection Attacks for Multi-Turn Conversations
Christian Schlarmann, Matthias Hein
Generative large vision-language models (LVLMs) have recently achieved impressive performance gains, and their user base is growing rapidly. However, the security of LVLMs, in part…
Unlearning That Lasts: Utility-Preserving, Robust, and Almost Irreversible Forgetting in LLMs
Naman Deep Singh, Maximilian Müller, Francesco Croce +1
Unlearning in large language models (LLMs) involves precisely removing specific information from a pre-trained model. This is crucial to ensure safety of LLMs by deleting private d…
Robustness in Both Domains: CLIP Needs a Robust Text Encoder
Elias Abad Rocamora, Christian Schlarmann, Naman Deep Singh +3
Adversarial input attacks can cause a significant shift of CLIP embeddings. This can affect the downstream robustness of models incorporating CLIP in the pipeline, such as text-to-…
FuseLIP: Multimodal Embeddings via Early Fusion of Discrete Tokens
Christian Schlarmann, Francesco Croce, Nicolas Flammarion +1
Contrastive language-image pre-training aligns features of text-image pairs in a common latent space via distinct encoders for each modality. While this approach achieves impressiv…
Adversarially Robust CLIP Models Can Induce Better (Robust) Perceptual Metrics
Francesco Croce, Christian Schlarmann, Naman Deep Singh +1
Measuring perceptual similarity is a key tool in computer vision. In recent years perceptual metrics based on features extracted from neural networks with large and diverse trainin…
Perturb and Recover: Fine-tuning for Effective Backdoor Removal from CLIP
Naman Deep Singh, Francesco Croce, Matthias Hein
Vision-Language models like CLIP have been shown to be highly effective at linking visual perception and natural language understanding, enabling sophisticated image-text capabilit…
Beyond Performance: Quantifying and Mitigating Label Bias in LLMs
Yuval Reif, Roy Schwartz
Large language models (LLMs) have shown remarkable adaptability to diverse tasks, by leveraging context prompts containing instructions, or minimal input-output examples. However,…
Robust CLIP: Unsupervised Adversarial Fine-Tuning of Vision Embeddings for Robust Large Vision-Language Models
Christian Schlarmann, Naman Deep Singh, Francesco Croce +1
Multi-modal foundation models like OpenFlamingo, LLaVA, and GPT-4 are increasingly used for various real-world tasks. Prior work has shown that these models are highly vulnerable t…
Segment (Almost) Nothing: Prompt-Agnostic Adversarial Attacks on Segmentation Models
Francesco Croce, Matthias Hein
General purpose segmentation models are able to generate (semantic) segmentation masks from a variety of prompts, including visual (points, boxed, etc.) and textual (object names)…
DiG-IN: Diffusion Guidance for Investigating Networks -- Uncovering Classifier Differences Neuron Visualisations and Visual Counterfactual Explanations
Maximilian Augustin, Yannic Neuhaus, Matthias Hein
While deep learning has led to huge progress in complex image classification tasks like ImageNet, unexpected failure modes, e.g. via spurious features, call into question how relia…