A survey on Concept-based Approaches For Model Improvement

TL;DR

This survey reviews concept-based explainability methods in deep neural networks, focusing on their role in interpretability, bias detection, and model improvement within vision tasks.

Contribution

It provides the first comprehensive taxonomy and systematic review of concept representations, discovery algorithms, and their application in model enhancement.

Findings

Various concept representation methods are categorized and analyzed.

Concept-based explanations help detect biases and spurious correlations.

Recent methods leverage concepts for improving model interpretability and generalization.

Abstract

The focus of recent research has shifted from merely improving the metrics based performance of Deep Neural Networks (DNNs) to DNNs which are more interpretable to humans. The field of eXplainable Artificial Intelligence (XAI) has observed various techniques, including saliency-based and concept-based approaches. These approaches explain the model's decisions in simple human understandable terms called Concepts. Concepts are known to be the thinking ground of humans}. Explanations in terms of concepts enable detecting spurious correlations, inherent biases, or clever-hans. With the advent of concept-based explanations, a range of concept representation methods and automatic concept discovery algorithms have been introduced. Some recent works also use concepts for model improvement in terms of interpretability and generalization. We provide a systematic review and taxonomy of various concept representations and their discovery algorithms in DNNs, specifically in vision. We also provide details on concept-based model improvement literature marking the first comprehensive survey of these methods.