Institution profile

Fraunhofer Heinrich Hertz Institute

Academic institutioneurope · de
Official website
Research library96linked papers
Opportunities0open roles
Selected work

Representative Papers

Manipulating Feature Visualizations with Gradient Slingshots

Jan 11, 2024arXiv.org

This work exposes a critical credibility vulnerability in feature visualization (FV) for deep neural network interpretability: FV outputs are susceptible to stealthy manipulation, leading to erroneous attribution of neuron semantics. To address this, we propose the first model-architecture-agnostic targeted FV manipulation method. Our approach integrates gradient redirection (via Slingshot optimization), adversarial latent-space perturbations, and neuron-activation-constrained regularization to achieve “semantic masking”—i.e., seamless substitution of a target neuron’s original FV explanation with an arbitrary user-specified semantic concept. Experiments across CNNs and Vision Transformers demonstrate successful concealment of functionally critical neurons: model accuracy degrades by less than 0.3%, yet FV-based auditing yields a 92% false-negative rate in detecting manipulated neurons. These results underscore the fragility of prevailing FV techniques and establish a new paradigm for robust model auditing and interpretability governance.

6 citationsRead paper

Spatially Resolved Meteorological and Ancillary Data in Central Europe for Rainfall Streamflow Modeling

Jun 04, 2025

The lack of high-resolution, spatially consistent hydrological input data for distributed modeling across Central Europe hinders the transition of neural network–based rainfall–runoff modeling from lumped to distributed frameworks. Method: This study constructs a comprehensive, daily-scale (1981–2011), 9 km × 9 km gridded hydrological dataset covering five major river basins, integrating multi-source heterogeneous geospatial data—including meteorological forcing, soil properties, lithology, land cover, and topography. Contribution/Results: It achieves, for the first time, standardized, spatially consistent gridding and fusion of hydrological data across multiple Central European basins. An open-source Python toolchain is developed, enabling seamless integration with observed streamflow records. The publicly released spatiotemporally aligned dataset—accompanied by reproducible code—significantly enhances both the generalizability and physical interpretability of deep learning–based hydrological models.

1 citationsRead paper
Recent publications

Latest Papers

What Does Animal Re-Identification Learn? Linear Biological Concepts and Their Origins in Visual Representations

Sep 05, 2026

Conservation increasingly relies on camera traps that collect more wildlife imagery than experts can manually analyze, making animal re-identification (Re-ID) essential for monitoring individuals and populations. Yet understanding which cues drive model decisions is challenging for ViT-based Re-ID models, whose metric-learning objectives provide no explicit supervision for biological concepts. We ask whether such models nonetheless organize their representations along biologically meaningful axes. Using a DINOv3 backbone fine-tuned for Western lowland gorilla Re-ID with triplet-margin loss, we find that sex and age emerge as linear directions that generalize to held-out individuals, reaching up to 0.91 AUROC and being recoverable from a single image per individual. Activation steering further shows that the sex direction is causally used by the model, flipping a significant fraction of predictions to the opposite sex. Comparing off-the-shelf and fine-tuned backbones shows that Re-ID training does not create these concepts, but relocates them across the network. Finally, data attribution reveals that the representation we find reflects a graded biological axis, is redundantly encoded across the population and shaped by visually ambiguous individuals. Together, these findings show how interpretability can uncover both the biological structure and failure modes of Re-ID representations, providing a step toward auditable computer vision for wildlife monitoring.

0 citationsRead paper