Improved Predictive Performance and Interpretability for Mesomorphic Neural Networks Using Local Fidelity Regularization

2026-06-29

AI Summary

This work addresses the unreliability of explanations in original Interpretable Mesoscopic Networks (IMNs), which lack constraints and are prone to degenerate solutions—such as weight collapse onto a single dimension. To mitigate this, the authors propose Local Fidelity Regularization (LFR), which enforces alignment between the weights of the linear output layer and local data variations, combined with an L1 sparsity penalty. This approach simultaneously prevents degenerate solutions, enhances predictive performance, and improves the trustworthiness of model explanations. On OpenML benchmark datasets, the regularized IMN significantly outperforms its unregularized counterpart in terms of AUROC, yields reliable interpretations, and achieves predictive accuracy comparable to state-of-the-art black-box models, thereby effectively overcoming the traditional trade-off between accuracy and interpretability.

Abstract

Interpretable Mesomorphic Neural Networks (IMNs) offer a promising framework that combines the predictive power of deep neural networks with the interpretability of linear models. However, the original formulation lacks safeguards to ensure that the learned interpretations are in fact reliable. In particular, the network is free to concentrate all explanatory variance into a single weight of the linear output layer, achieving strong predictive performance while producing interpretations that are largely meaningless. Paradoxically, the L1 penalty proposed to encourage sparse solutions exacerbates this problem by further incentivizing such degenerate configurations. To address this vulnerability, we introduce Local Fidelity Regularization (LFR), a novel penalty term that prevents degenerate weight collapse by aligning the linear output weights with local data variations. This structural constraint guarantees faithful explanations and substantially improves the reliability of model interpretations. Furthermore, empirical evaluations across the OpenML benchmark suite demonstrate that LFR does not compromise accuracy for explainability; rather, it achieved improved AUROC over the unregularized IMN. By yielding results highly competitive with state-of-the-art black-box models, LFR provides the dual benefit of reliable interpretability and superior predictive performance. Source code and usage instructions are available at https://github.com/hugohammer/LFR-IMN.git.

Problem

• Interpretable Mesomorphic Neural Networks
• Local Fidelity Regularization
• degenerate weight collapse
• reliable interpretability
• explanatory variance

Innovation

• Local Fidelity Regularization
• Interpretable Mesomorphic Neural Networks
• Explainable AI
• Weight Collapse Prevention
• Predictive Performance