Formal Concept Lattices are Good Semantic Scaffolds for Concept-Based Learning
For researchers in interpretable AI, this work provides a principled way to align neural network hierarchies with explicit semantic structures, though it is an incremental improvement over existing concept-based models.
The paper introduces a method using formal concept lattices to impose hierarchical semantic structure on concept-based neural networks, resulting in more interpretable embeddings and effective interventions.
Learning semantics is essential for deep learning models to be interpretable and better aligned with human reasoning. Concept-based models approach this by representing classes through meaningful semantic abstractions, but typically treat all concepts as a flat, unstructured set learned at a single neural network layer. This overlooks a fundamental property of human semantic understanding: concepts being organized hierarchically, from general to specific. While deep networks do learn a hierarchy of visual features, this structure is rarely aligned with explicit semantic hierarchies. Drawing on Formal Concept Analysis, we demonstrate that formal concept lattices provide principled semantic scaffolds to guide neural network learning. These lattices naturally identify where in the network concepts should be learned based on their level of generality. This allows the model to develop staged, semantically grounded representations throughout its depth. Empirical results on real-world datasets show that our models produce more interpretable embeddings, support more effective interventions, and learn concept representations that are both meaningful and hierarchically structured.