7 papers · 1 filter
Interpretability-Guided Soft Pruning of Attention Heads in Vision Transformers
Kamil KsiÄ Å¼ek, Piotr SuszyÅski, MichaÅ Jan WÅodarczyk +2
Vision foundation models, such as DINOv2, learn highly expressive representations but rely on massive, opaque architectures that demand substantial computational power and memory.…
What Pixels Are Enough? SEAMS: Sufficiency Saliency via MSE-Preservation Soft-Masks
Magdalena TrÄdowicz, Åukasz Struski, Arkadiusz Lewicki +4
Saliency maps are most useful when they identify the image regions that are sufficient to preserve a model's behaviour. We introduce SEAMS, a sufficiency-based saliency method that…
Conceptualizing Embeddings: Sparse Disentanglement for Vision-Language Models
Piotr Kubaty, Patryk MarszaÅek, Åukasz Struski +3
Vision-language models learn powerful multimodal embeddings, yet their internal semantics remain opaque. While sparse autoencoders (SAEs) can extract interpretable features, they r…
DAVE: Distribution-aware Attribution via ViT Gradient Decomposition
Adam Wróbel, Siddhartha Gairola, Jacek Tabor +3
Vision Transformers (ViTs) have become a dominant architecture in computer vision, yet producing stable and high-resolution attribution maps for these models remains challenging. A…
SIDE: Sparse Information Disentanglement for Explainable Artificial Intelligence
Viktar Dubovik, Åukasz Struski, Jacek Tabor +1
Understanding the decisions made by deep neural networks is essential in high-stakes domains such as medical imaging and autonomous driving. Yet, these models often lack transparen…
InfoDisent: Explainability of Image Classification Models by Information Disentanglement
Åukasz Struski, Dawid Rymarczyk, Jacek Tabor
In this work, we introduce InfoDisent, a hybrid approach to explainability based on the information bottleneck principle. InfoDisent enables the disentanglement of information in t…