Counterfactual Explanation
Counterfactual explanations (CFEs) aim to enhance the interpretability of machine learning models by showing how minimal input changes would alter predictions. Current research focuses on developing robust and efficient CFE generation methods across various model types, including deep learning architectures like variational autoencoders and diffusion models, and for diverse data modalities such as images, time series, and text. This work is significant because CFEs improve model transparency and trustworthiness, fostering greater user understanding and facilitating the responsible deployment of AI in high-stakes applications like healthcare and finance.
Papers
TACE: Tumor-Aware Counterfactual Explanations
Eleonora Beatrice Rossi, Eleonora Lopez, Danilo Comminiello
The Gaussian Discriminant Variational Autoencoder (GdVAE): A Self-Explainable Model with Counterfactual Explanations
Anselm Haselhoff, Kevin Trelenberg, Fabian Küppers, Jonas Schneider
Counterfactual Explanations for Clustering Models
Aurora Spagnol, Kacper Sokol, Pietro Barbiero, Marc Langheinrich, Martin Gjoreski
Evaluating the Reliability of Self-Explanations in Large Language Models
Korbinian Randl, John Pavlopoulos, Aron Henriksson, Tony Lindgren
Contrastive Learning with Counterfactual Explanations for Radiology Report Generation
Mingjie Li, Haokun Lin, Liang Qiu, Xiaodan Liang, Ling Chen, Abdulmotaleb Elsaddik, Xiaojun Chang