Zobrazeno 1 - 10
of 510
pro vyhledávání: '"Hein, Matthias A."'
A plethora of jailbreaking attacks have been proposed to obtain harmful responses from safety-tuned LLMs. In their original settings, these methods all largely succeed in coercing the target output, but their attacks vary substantially in fluency and
Externí odkaz:
http://arxiv.org/abs/2410.16222
Autor:
Mueller, Maximilian, Hein, Matthias
In realistic medical settings, the data are often inherently long-tailed, with most samples concentrated in a few classes and a long tail of rare classes, usually containing just a few samples. This distribution presents a significant challenge becau
Externí odkaz:
http://arxiv.org/abs/2409.01317
Inverse problems, such as accelerated MRI reconstruction, are ill-posed and an infinite amount of possible and plausible solutions exist. This may not only lead to uncertainty in the reconstructed image but also in downstream tasks such as semantic s
Externí odkaz:
http://arxiv.org/abs/2407.18026
Autor:
Andrich, Carsten, Nowack, Tobias F., Ihlow, Alexander, Giehl, Sebastian, Engelhardt, Maximilian, Sommerkorn, Gerd, Schwind, Andreas, Hofmann, Willi, Bornkessel, Christian, Thomä, Reiner S., Hein, Matthias A.
The upcoming 6G mobile communication standard will offer a revolutionary new feature: Integrated sensing and communication (ISAC) reuses mobile communication signals to realize multi-static radar for various applications including localization. Conse
Externí odkaz:
http://arxiv.org/abs/2407.13749
Autor:
Peleg, Amit, Hein, Matthias
Neural networks typically generalize well when fitting the data perfectly, even though they are heavily overparameterized. Many factors have been pointed out as the reason for this phenomenon, including an implicit bias of stochastic gradient descent
Externí odkaz:
http://arxiv.org/abs/2407.03848
Autor:
Mueller, Maximilian, Hein, Matthias
VisionTransformers have been shown to be powerful out-of-distribution detectors for ImageNet-scale settings when finetuned from publicly available checkpoints, often outperforming other model types on popular benchmarks. In this work, we investigate
Externí odkaz:
http://arxiv.org/abs/2405.17447
Multi-modal foundation models such as CLIP have showcased impressive zero-shot capabilities. However, their applicability in resource-constrained environments is limited due to their large number of parameters and high inference time. While existing
Externí odkaz:
http://arxiv.org/abs/2404.16637
Many safety-critical applications, especially in autonomous driving, require reliable object detectors. They can be very effectively assisted by a method to search for and identify potential failures and systematic errors before these detectors are d
Externí odkaz:
http://arxiv.org/abs/2404.07045
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 to adversarial attacks on the vision modality. These attacks can be lev
Externí odkaz:
http://arxiv.org/abs/2402.12336
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 reliably these classifiers work in the wild. Furthermore, for safety-critic
Externí odkaz:
http://arxiv.org/abs/2311.17833