1k citations · 4.2k across the 120 of their papers we have counts for
130 papers
Reliability Scaling Laws for Quantized Large Language Models
Sirine Ayadi, Sándor Daróczi, Stephan Günnemann +1
Quantization is a powerful strategy to build capable and resource-efficient large language models (LLMs) by reducing the bitwidth of the parameters. While quantized LLMs achieve st…
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…
Speculative Sampling For Faster Molecular Dynamics
Arthur Kosmala, Stephan Günnemann, Meng Gao +1
Molecular dynamics (MD) is a key tool for simulating the dynamical behavior of atomic systems. However, MD is inherently serial, which makes it difficult to increase single-system…
Provable Robustness against Backdoor Attacks via the Primal-Dual Perspective on Differential Privacy
Aman Saxena, Jan Schuchardt, Yan Scholten +1
Randomized smoothing is a powerful tool for certifying robustness to adversarial perturbations, including poisoning attacks via randomized training and evasion attacks via randomiz…
Interpolating Discrete Diffusion Models with Controllable Resampling
Marcel Kollovieh, Sirine Ayadi, Stephan Günnemann
Discrete diffusion models form a powerful class of generative models across diverse domains, including text and graphs. However, existing approaches face fundamental limitations. M…
Exact Certification of Neural Networks and Partition Aggregation Ensembles against Label Poisoning
Ajinkya Mohgaonkar, Lukas Gosch, Mahalakshmi Sabanayagam +2
Label-flipping attacks, which corrupt training labels to induce misclassifications at inference, remain a major threat to supervised learning models. This drives the need for robus…
Amplified Patch-Level Differential Privacy for Free via Random Cropping
Kaan Durmaz, Jan Schuchardt, Sebastian Schmidt +1
Random cropping is one of the most common data augmentation techniques in computer vision, yet the role of its inherent randomness in training differentially private machine learni…
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…
Flow-Based Density Ratio Estimation for Intractable Distributions with Applications in Genomics
Egor Antipov, Alessandro Palma, Lorenzo Consoli +3
Estimating density ratios between pairs of intractable data distributions is a core problem in probabilistic modeling, enabling principled comparisons of sample likelihoods under d…
3D Molecule Generation from Rigid Motifs via SE(3) Flows
Roman Poletukhin, Marcel Kollovieh, Eike Eberhard +1
Three-dimensional molecular structure generation is typically performed at the level of individual atoms, yet molecular graph generation techniques often consider fragments as thei…
Task-Awareness Improves LLM Generations and Uncertainty
Tim Tomov, Dominik Fuchsgruber, Stephan Günnemann
In many applications of LLMs, natural language responses often have an underlying structure such as representing discrete labels, numerical values, or graphs. Yet, existing decodin…
Modeling Microenvironment Trajectories on Spatial Transcriptomics with NicheFlow
Kristiyan Sakalyan, Alessandro Palma, Filippo Guerranti +2
Understanding the evolution of cellular microenvironments in spatiotemporal data is essential for deciphering tissue development and disease progression. While experimental techniq…
Unexplored flaws in multiple-choice VQA make benchmarking unreliable
Fabio Rosenthal, Sebastian Schmidt, Thorsten Graf +3
Previous works identify sensitivity to option order as a key issue in multiple-choice VQA (MC-VQA) evaluation and propose protocols to mitigate this effect. We show that such mitig…
Discrete Bayesian Sample Inference for Graph Generation
Ole Petersen, Marcel Kollovieh, Marten Lienen +1
Generating graph-structured data is crucial in applications such as molecular generation, knowledge graphs, and network analysis. However, their discrete, unordered nature makes th…
What Expressivity Theory Misses: Message Passing Complexity for GNNs
Niklas Kemper, Tom Wollschläger, Stephan Günnemann
Expressivity theory, characterizing which graphs a GNN can distinguish, has become the predominant framework for analyzing GNNs, with new models striving for higher expressivity. H…
A Machine Learning Perspective on Automated Driving Corner Cases
Sebastian Schmidt, Julius Körner, Stephan Günnemann
For high-stakes applications, like autonomous driving, a safe operation is necessary to prevent harm, accidents, and failures. Traditionally, difficult scenarios have been categori…
Edit-Based Flow Matching for Temporal Point Processes
David Lüdke, Marten Lienen, Marcel Kollovieh +1
Temporal point processes (TPPs) are a fundamental tool for modeling event sequences in continuous time, but most existing approaches rely on autoregressive parameterizations that a…
Predictive Feature Caching for Training-free Acceleration of Molecular Geometry Generation
Johanna Sommer, John Rachwan, Nils Fleischmann +2
Flow matching models generate high-fidelity molecular geometries but incur significant computational costs during inference, requiring hundreds of network evaluations. This inferen…
Energy-Weighted Flow Matching: Unlocking Continuous Normalizing Flows for Efficient and Scalable Boltzmann Sampling
Niclas Dern, Lennart Redl, Sebastian Pfister +3
Sampling from unnormalized target distributions, e.g.\ Boltzmann distributions , is fundamental to many scientific applications yet comp…
Enforcing Latent Euclidean Geometry in Single-Cell VAEs for Manifold Interpolation
Alessandro Palma, Sergei Rybakov, Leon Hetzel +2
Latent space interpolations are a powerful tool for navigating deep generative models in applied settings. An example is single-cell RNA sequencing, where existing methods model ce…
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…
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…