10 papers
Steering Fields: Adaptive Vector Fields for Safe Image Generation and Beyond
Simone Facchiano, Jan Eric Lenssen, Bernt Schiele +3
As state-of-the-art text-to-image flow models achieve near-photorealistic quality, controlling their outputs, e.g., suppressing harmful content while promoting benign alternatives,…
Interpretability Without Tradeoffs: Disentangling Polysemanticity At Equal Predictive Performance
Doğukan Bağcı, Bernt Schiele, Simone Schaub-Meyer +2
Deep neural networks (DNNs) are widely used, but interpreting what they actually learn remains difficult. A major obstacle is that individual neurons often encode multiple unrelate…
Seeing Through Circuits: Faithful Mechanistic Interpretability for Vision Transformers
Nina Żukowska, Wolfgang Stammer, Bernt Schiele +1
Transparency of neural networks' internal reasoning is at the heart of interpretability research, adding to trust, safety, and understanding of these models. The field of mechanist…
Certified Circuits: Stability Guarantees for Mechanistic Circuits
Alaa Anani, Tobias Lorenz, Bernt Schiele +2
Understanding how neural networks arrive at their predictions is essential for debugging, auditing, and deployment. Mechanistic interpretability pursues this goal by identifying ci…
CFM: Language-aligned Concept Foundation Model for Vision
Kai Wittenmayer, Sukrut Rao, Amin Parchami-Araghi +2
Language-aligned vision foundation models perform strongly across diverse downstream tasks. Yet, their learned representations remain opaque, making interpreting their decision-mak…
FaCT: Faithful Concept Traces for Explaining Neural Network Decisions
Amin Parchami-Araghi, Sukrut Rao, Jonas Fischer +1
Deep networks have shown remarkable performance across a wide range of tasks, yet getting a global concept-level understanding of how they function remains a key challenge. Many po…
LayerCake: Token-Aware Contrastive Decoding within Large Language Model Layers
Jingze Zhu, Yongliang Wu, Wenbo Zhu +7
Large language models (LLMs) excel at natural language understanding and generation but remain vulnerable to factual errors, limiting their reliability in knowledge-intensive tasks…
Disentangling Polysemantic Channels in Convolutional Neural Networks
Robin Hesse, Jonas Fischer, Simone Schaub-Meyer +1
Mechanistic interpretability is concerned with analyzing individual components in a (convolutional) neural network (CNN) and how they form larger circuits representing decision mec…
VITAL: More Understandable Feature Visualization through Distribution Alignment and Relevant Information Flow
Ada Gorgun, Bernt Schiele, Jonas Fischer
Neural networks are widely adopted to solve complex and challenging tasks. Especially in high-stakes decision-making, understanding their reasoning process is crucial, yet proves c…
Interpretable 3D Neural Object Volumes for Robust Conceptual Reasoning
Nhi Pham, Artur Jesslen, Bernt Schiele +2
With the rise of deep neural networks, especially in safety-critical applications, robustness and interpretability are crucial to ensure their trustworthiness. Recent advances in 3…