Sparse Linear Concept Discovery Models

TL;DR

This paper introduces a novel interpretable model combining Contrastive Language Image models with a sparse linear layer, achieving superior accuracy and high concept sparsity for better interpretability in decision-making.

Contribution

It presents a Bayesian-based sparse linear framework that improves accuracy and interpretability over existing Concept Bottleneck Models.

Findings

Outperforms recent CBM approaches in accuracy.

Achieves high per-example concept sparsity.

Facilitates individual investigation of concepts.

Abstract

The recent mass adoption of DNNs, even in safety-critical scenarios, has shifted the focus of the research community towards the creation of inherently intrepretable models. Concept Bottleneck Models (CBMs) constitute a popular approach where hidden layers are tied to human understandable concepts allowing for investigation and correction of the network's decisions. However, CBMs usually suffer from: (i) performance degradation and (ii) lower interpretability than intended due to the sheer amount of concepts contributing to each decision. In this work, we propose a simple yet highly intuitive interpretable framework based on Contrastive Language Image models and a single sparse linear layer. In stark contrast to related approaches, the sparsity in our framework is achieved via principled Bayesian arguments by inferring concept presence via a data-driven Bernoulli distribution. As we experimentally show, our framework not only outperforms recent CBM approaches accuracy-wise, but it also yields high per example concept sparsity, facilitating the individual investigation of the emerging concepts.