HINT: Hierarchical Neuron Concept Explainer
This work addresses the need for more structured and scalable interpretability methods in deep learning, though it is incremental as it builds on existing concept-based approaches.
The paper tackles the problem of interpreting deep networks by associating neurons with hierarchical concepts, proposing HINT to build bidirectional associations between neurons and concepts at different semantic levels, and demonstrates its applicability in tasks like discovering saliency regions and explaining adversarial attacks.
To interpret deep networks, one main approach is to associate neurons with human-understandable concepts. However, existing methods often ignore the inherent relationships of different concepts (e.g., dog and cat both belong to animals), and thus lose the chance to explain neurons responsible for higher-level concepts (e.g., animal). In this paper, we study hierarchical concepts inspired by the hierarchical cognition process of human beings. To this end, we propose HIerarchical Neuron concepT explainer (HINT) to effectively build bidirectional associations between neurons and hierarchical concepts in a low-cost and scalable manner. HINT enables us to systematically and quantitatively study whether and how the implicit hierarchical relationships of concepts are embedded into neurons, such as identifying collaborative neurons responsible to one concept and multimodal neurons for different concepts, at different semantic levels from concrete concepts (e.g., dog) to more abstract ones (e.g., animal). Finally, we verify the faithfulness of the associations using Weakly Supervised Object Localization, and demonstrate its applicability in various tasks such as discovering saliency regions and explaining adversarial attacks. Code is available on https://github.com/AntonotnaWang/HINT.