8 papers
A Granular Study of Safety Pretraining under Model Abliteration
Shashank Agnihotri, Jonas Jakubassa, Priyam Dey +4
Open-weight LLMs can be modified at inference time with simple activation edits, which raises a practical question for safety: do common safety interventions like refusal training…
-Quant: Towards Learnable Quantization for Low-bit Pattern Recognition
Mishal Fatima, Shashank Agnihotri, Marius Bock +4
Most pattern recognition models are developed on pre-proce\-ssed data. In computer vision, for instance, RGB images processed through image signal processing (ISP) pipelines design…
AIM: Amending Inherent Interpretability via Self-Supervised Masking
Eyad Alshami, Shashank Agnihotri, Bernt Schiele +1
It has been observed that deep neural networks (DNNs) often use both genuine as well as spurious features. In this work, we propose "Amending Inherent Interpretability via Self-Sup…
Faithful, Interpretable Chest X-ray Diagnosis with Anti-Aliased B-cos Networks
Marcel Kleinmann, Shashank Agnihotri, Margret Keuper
Faithfulness and interpretability are essential for deploying deep neural networks (DNNs) in safety-critical domains such as medical imaging. B-cos networks offer a promising solut…
Smart Eyes for Silent Threats: VLMs and In-Context Learning for THz Imaging
Nicolas Poggi, Shashank Agnihotri, Margret Keuper
Terahertz (THz) imaging enables non-invasive analysis for applications such as security screening and material classification, but effective image classification remains challengin…
SemSegBench & DetecBench: Benchmarking Reliability and Generalization Beyond Classification
Shashank Agnihotri, David Schader, Jonas Jakubassa +5
Reliability and generalization in deep learning are predominantly studied in the context of image classification. Yet, real-world applications in safety-critical domains involve a…