Saved in:
Bibliographic Details
Main Authors: Zeid, Ahmed, Bender, Sidney
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.17048
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910057533276160
author Zeid, Ahmed
Bender, Sidney
author_facet Zeid, Ahmed
Bender, Sidney
contents Modern neural networks achieve strong performance but remain difficult to interpret in high-dimensional visual domains. Counterfactual explanations (CFEs) provide a principled approach to interpreting black-box predictions by identifying minimal input changes that alter model outputs. However, existing CFE methods often rely on dataset-specific generative models and incur substantial computational cost, limiting their scalability to high-resolution data. We propose SCE-LITE-HQ, a scalable framework for counterfactual generation that leverages pretrained generative foundation models without task-specific retraining. The method operates in the latent space of the generator, incorporates smoothed gradients to improve optimization stability, and applies mask-based diversification to promote realistic and structurally diverse counterfactuals. We evaluate SCE-LITE-HQ on natural and medical datasets using a desiderata-driven evaluation protocol. Results show that SCE-LITE-HQ produces valid, realistic, and diverse counterfactuals competitive with or outperforming existing baselines, while avoiding the overhead of training dedicated generative models.
format Preprint
id arxiv_https___arxiv_org_abs_2603_17048
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SCE-LITE-HQ: Smooth visual counterfactual explanations with generative foundation models
Zeid, Ahmed
Bender, Sidney
Machine Learning
Computer Vision and Pattern Recognition
Modern neural networks achieve strong performance but remain difficult to interpret in high-dimensional visual domains. Counterfactual explanations (CFEs) provide a principled approach to interpreting black-box predictions by identifying minimal input changes that alter model outputs. However, existing CFE methods often rely on dataset-specific generative models and incur substantial computational cost, limiting their scalability to high-resolution data. We propose SCE-LITE-HQ, a scalable framework for counterfactual generation that leverages pretrained generative foundation models without task-specific retraining. The method operates in the latent space of the generator, incorporates smoothed gradients to improve optimization stability, and applies mask-based diversification to promote realistic and structurally diverse counterfactuals. We evaluate SCE-LITE-HQ on natural and medical datasets using a desiderata-driven evaluation protocol. Results show that SCE-LITE-HQ produces valid, realistic, and diverse counterfactuals competitive with or outperforming existing baselines, while avoiding the overhead of training dedicated generative models.
title SCE-LITE-HQ: Smooth visual counterfactual explanations with generative foundation models
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.17048