Provable concept learning for interpretable predictions using variational autoencoders
This addresses the need for interpretable AI in safety-critical domains like healthcare, offering a novel approach to concept learning.
The paper tackles the problem of neural network interpretability in safety-critical applications by proposing a VAE-based classifier that identifies high-level, previously unknown ground-truth concepts for predictions, proving it can identify these concepts while achieving optimal classification accuracy and showing promising results on a medical dataset.
In safety-critical applications, practitioners are reluctant to trust neural networks when no interpretable explanations are available. Many attempts to provide such explanations revolve around pixel-based attributions or use previously known concepts. In this paper we aim to provide explanations by provably identifying \emph{high-level, previously unknown ground-truth concepts}. To this end, we propose a probabilistic modeling framework to derive (C)oncept (L)earning and (P)rediction (CLAP) -- a VAE-based classifier that uses visually interpretable concepts as predictors for a simple classifier. Assuming a generative model for the ground-truth concepts, we prove that CLAP is able to identify them while attaining optimal classification accuracy. Our experiments on synthetic datasets verify that CLAP identifies distinct ground-truth concepts on synthetic datasets and yields promising results on the medical Chest X-Ray dataset.