DifCluE: Generating Counterfactual Explanations with Diffusion Autoencoders and modal clustering

TL;DR

DifCluE leverages diffusion autoencoders and modal clustering to generate diverse, reliable counterfactual explanations, improving interpretability by capturing multiple modes within classes.

Contribution

This paper introduces DifCluE, a novel method that uses diffusion autoencoders and latent space clustering to produce multiple meaningful counterfactuals for complex data distributions.

Findings

Outperforms state-of-the-art in counterfactual generation

Effectively uncovers multiple data modes within classes

Provides more reliable and diverse explanations

Abstract

Generating multiple counterfactual explanations for different modes within a class presents a significant challenge, as these modes are distinct yet converge under the same classification. Diffusion probabilistic models (DPMs) have demonstrated a strong ability to capture the underlying modes of data distributions. In this paper, we harness the power of a Diffusion Autoencoder to generate multiple distinct counterfactual explanations. By clustering in the latent space, we uncover the directions corresponding to the different modes within a class, enabling the generation of diverse and meaningful counterfactuals. We introduce a novel methodology, DifCluE, which consistently identifies these modes and produces more reliable counterfactual explanations. Our experimental results demonstrate that DifCluE outperforms the current state-of-the-art in generating multiple counterfactual explanations, offering a significant advancement in model interpretability.