27 citations · 110 across the 41 of their papers we have counts for
45 papers
Visual Memory Attacks Can Persist Through The KV Cache
David Dobre, Leo Schwinn, Gauthier Gidel +3
Modern language model systems operate autonomously over increasingly long contexts containing untrusted text and images. Can an adversarial input continue to steer a model even aft…
All you need to break LLMs are Black-Box, Adapting, Efficient, Transferable, Harmful, Applicable ... Attacks
Vincent Limbach, Jonas Dornbusch, David Lüdke +2
Accurately evaluating adversarial robustness is a longstanding challenge. A flawed attack design can inflate robustness estimates, making deployment risk assessment and defense com…
A Coin Flip for Safety: LLM Judges Fail to Reliably Measure Adversarial Robustness
Leo Schwinn, Moritz Ladenburger, Tim Beyer +3
Automated \enquote{LLM-as-a-Judge} frameworks have become the de facto standard for scalable evaluation across natural language processing. For instance, in safety evaluation, thes…
Closing the Distribution Gap in Adversarial Training for LLMs
Chengzhi Hu, Jonas Dornbusch, David Lüdke +2
Adversarial training for LLMs is one of the most promising methods to reliably improve robustness against adversaries. However, despite significant progress, models remain vulnerab…
AdversariaLLM: A Unified and Modular Toolbox for LLM Robustness Research
Tim Beyer, Jonas Dornbusch, Jakob Steimle +3
The rapid expansion of research on Large Language Model (LLM) safety and robustness has produced a fragmented and oftentimes buggy ecosystem of implementations, datasets, and evalu…
Diffusion LLMs are Natural Adversaries for any LLM
David Lüdke, Tom Wollschläger, Paul Ungermann +2
We introduce a novel framework that transforms the resource-intensive (adversarial) prompt optimization problem into an \emph{efficient, amortized inference task}. Our core insight…
Sampling-aware Adversarial Attacks Against Large Language Models
Tim Beyer, Yan Scholten, Leo Schwinn +1
To guarantee safe and robust deployment of large language models (LLMs) at scale, it is critical to accurately assess their adversarial robustness. Existing adversarial attacks typ…
Model Collapse Is Not a Bug but a Feature in Machine Unlearning for LLMs
Yan Scholten, Sophie Xhonneux, Leo Schwinn +1
Current unlearning methods for LLMs optimize on the private information they seek to remove by incorporating it into their fine-tuning data. We argue this not only risks reinforcin…
Effective Data Pruning through Score Extrapolation
Sebastian Schmidt, Prasanga Dhungel, Christoffer Löffler +3
Training advanced machine learning models demands massive datasets, resulting in prohibitive computational costs. To address this challenge, data pruning techniques identify and re…
Jailbreak Distillation: Renewable Safety Benchmarking
Jingyu Zhang, Ahmed Elgohary, Xiawei Wang +5
Large language models (LLMs) are rapidly deployed in critical applications, raising urgent needs for robust safety benchmarking. We propose Jailbreak Distillation (JBDistill), a no…
When to retrain a machine learning model
Regol Florence, Schwinn Leo, Sprague Kyle +2
A significant challenge in maintaining real-world machine learning models is responding to the continuous and unpredictable evolution of data. Most practitioners are faced with the…
Joint Relational Database Generation via Graph-Conditional Diffusion Models
Mohamed Amine Ketata, David Lüdke, Leo Schwinn +1
Building generative models for relational databases (RDBs) is important for many applications, such as privacy-preserving data release and augmenting real datasets. However, most p…
Byte Pair Encoding for Efficient Time Series Forecasting
Leon Götz, Marcel Kollovieh, Stephan Günnemann +1
Existing time series tokenization methods predominantly encode a constant number of samples into individual tokens. This inflexible approach can generate excessive tokens for even…
Understanding Cross-Model Perceptual Invariances Through Ensemble Metamers
Lukas Boehm, Jonas Leo Mueller, Christoffer Loeffler +3
Understanding the perceptual invariances of artificial neural networks is essential for improving explainability and aligning models with human vision. Metamers - stimuli that are…
Joint Out-of-Distribution Filtering and Data Discovery Active Learning
Sebastian Schmidt, Leonard Schenk, Leo Schwinn +1
As the data demand for deep learning models increases, active learning (AL) becomes essential to strategically select samples for labeling, which maximizes data efficiency and redu…
LLM-Safety Evaluations Lack Robustness
Tim Beyer, Sophie Xhonneux, Simon Geisler +3
In this paper, we argue that current safety alignment research efforts for large language models are hindered by many intertwined sources of noise, such as small datasets, methodol…
Adversarial Alignment for LLMs Requires Simpler, Reproducible, and More Measurable Objectives
Leo Schwinn, Yan Scholten, Tom Wollschläger +4
Misaligned research objectives have considerably hindered progress in adversarial robustness research over the past decade. For instance, an extensive focus on optimizing target me…
Fast Proxies for LLM Robustness Evaluation
Tim Beyer, Jan Schuchardt, Leo Schwinn +1
Evaluating the robustness of LLMs to adversarial attacks is crucial for safe deployment, yet current red-teaming methods are often prohibitively expensive. We compare the ability o…
A Generative Approach to LLM Harmfulness Mitigation with Red Flag Tokens
David Dobre, Mehrnaz Mofakhami, Sophie Xhonneux +2
Many safety post-training methods for large language models (LLMs) are designed to modify the model's behaviour from producing unsafe answers to issuing refusals. However, such dis…
Extracting Unlearned Information from LLMs with Activation Steering
Atakan Seyitoğlu, Aleksei Kuvshinov, Leo Schwinn +1
An unintended consequence of the vast pretraining of Large Language Models (LLMs) is the verbatim memorization of fragments of their training data, which may contain sensitive or c…
A Probabilistic Perspective on Unlearning and Alignment for Large Language Models
Yan Scholten, Stephan Günnemann, Leo Schwinn
Comprehensive evaluation of Large Language Models (LLMs) is an open research problem. Existing evaluations rely on deterministic point estimates generated via greedy decoding. Howe…
Flow Matching with Gaussian Process Priors for Probabilistic Time Series Forecasting
Marcel Kollovieh, Marten Lienen, David Lüdke +2
Recent advancements in generative modeling, particularly diffusion models, have opened new directions for time series modeling, achieving state-of-the-art performance in forecastin…
Caption-Driven Explorations: Aligning Image and Text Embeddings through Human-Inspired Foveated Vision
Dario Zanca, Andrea Zugarini, Simon Dietz +4
Understanding human attention is crucial for vision science and AI. While many models exist for free-viewing, less is known about task-driven image exploration. To address this, we…
Revisiting the Robust Alignment of Circuit Breakers
Leo Schwinn, Simon Geisler
Over the past decade, adversarial training has emerged as one of the few reliable methods for enhancing model robustness against adversarial attacks [Szegedy et al., 2014, Madry et…