# Faithful, Interpretable Chest X-ray Diagnosis with Artifact-free B-cos Networks

arXiv: https://arxiv.org/abs/2507.16761 · HTML: https://mlsanity.com/paper/2507.16761/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- First posted on arXiv 22 Jul 2025.

## Abstract (arXiv)

Faithfulness and interpretability are essential for deploying deep neural networks (DNNs) in safety-critical domains such as medical image analysis. B-cos networks modify the parameterization of convolutional and classification layers to measure class evidence via feature-weight alignment, enabling built-in, class-specific contribution maps without post-hoc explanations. While maintaining diagnostic performance competitive with state-of-the-art DNNs, standard B-cos networks exhibit severe aliasing artifacts in their explanation maps, rendering them unsuitable for clinical use, where clarity is essential. In this work, we address this limitation by introducing anti-aliasing strategies using ASAP and BlurPool (BP) to significantly improve explanation quality. Our experiments on chest X-ray datasets demonstrate that the modified B-cosASAP and B-cosBP preserve strong predictive performance while providing faithful and artifact-free explanations suitable for clinical application in multi-class and multi-label settings. Code is available at: https://github.com/shrebox/Artifact-free-B-cos-Networks.