Variational Autoencoders (VAEs) have shown strong performance in tasks such as out-of-distribution detection and image reconstruction, yet they have rarely been explored for concept extraction. Using a Sparse VAE could allow concepts to be modeled as multivariate Gaussian distributions, potentially offering new insights into model uncertainty. This project will investigate whether a Sparse VAE is suitable for concept extraction and, if successful, explore how it can be used to explain model uncertainty.

Literature

1. Fel, Thomas, et al. “Craft: Concept recursive activation factorization for explainability.”
2. Kingma, Diederik, Welling, Max “An Introduction to Variational Autoencoders”
3. Geadah, Victor, et al. “Sparse-Coding Variational Autoencoders”