8 papers
Structuring Sparsity: Block-Sparse Featurizers Capture Visual Concept Manifolds
Thomas Fel, Matthew Kowal, Mozes Jacobs +22
What is the geometry of a visual percept? The most widely used protocols for decomposing neural network representations into interpretable parts treat concepts as isolated directio…
A Unifying Framework for Concept-Based Representational Similarity
Grégoire Dhimoïla, Victor Boutin, Agustin Martin Picard +2
Learned representations across models and modalities often exhibit striking structural similarities, suggesting shared underlying concept decompositions. However, concept alignment…
Choosing the right basis for interpretability: Psychophysical comparison between neuron-based and dictionary-based representations
Julien Colin, Lore Goetschalckx, Thomas Fel +3
Interpretability research often adopts a neuron-centric lens, treating individual neurons as the fundamental units of explanation. However, neuron-level explanations can be undermi…
Cross-Modal Redundancy and the Geometry of Vision-Language Embeddings
Grégoire Dhimoïla, Thomas Fel, Victor Boutin +1
Vision-language models (VLMs) align images and text with remarkable success, yet the geometry of their shared embedding space remains poorly understood. To probe this geometry, we…
Back to the Baseline: Examining Baseline Effects on Explainability Metrics
Agustin Martin Picard, Thibaut Boissin, Varshini Subhash +2
Attribution methods are among the most prevalent techniques in Explainable Artificial Intelligence (XAI) and are usually evaluated and compared using Fidelity metrics, with Inserti…
One Wave To Explain Them All: A Unifying Perspective On Feature Attribution
Gabriel Kasmi, Amandine Brunetto, Thomas Fel +1
Feature attribution methods aim to improve the transparency of deep neural networks by identifying the input features that influence a model's decision. Pixel-based heatmaps have b…