A Rate-Distortion Framework for Explaining Black-box Model Decisions

2021/10/12 by Stefan Kolek, Đức Anh Nguyễn, Kolek, Stefan +7 · 1 citation
Computer Science · #Anomaly Detection Techniques and Applications #Artificial Intelligence (cs.AI) #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis #Information Theory (cs.IT) #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.2110.08252

openalex publication_date 2021/10/12 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We present the Rate-Distortion Explanation (RDE) framework, a mathematically well-founded method for explaining black-box model decisions. The framework is based on perturbations of the target input signal and applies to any differentiable pre-trained model such as neural networks. Our experiments demonstrate the framework's adaptability to diverse data modalities, particularly images, audio, and physical simulations of urban environments.

Cited by

• Direct Preference Optimization for Adaptive Concept-based Explanations

Related

• In-Distribution Interpretability for Challenging Modalities
• A Rate-Distortion Framework for Explaining Neural Network Decisions
• "Why did you do that?": Explaining black box models with Inductive Synthesis
• Shedding Light on Black Box Machine Learning Algorithms: Development of an Axiomatic Framework to Assess the Quality of Methods that Explain Individual Predictions
• What made you do this? Understanding black-box decisions with sufficient\n input subsets