Interpretable Geometric Deep Learning via Learnable Randomness Injection

TL;DR

This paper introduces learnable randomness injection (LRI), a novel mechanism for creating interpretable geometric deep learning models that identify scientifically meaningful points in point cloud data, improving robustness and interpretability.

Contribution

The work proposes LRI, a general approach for building inherently interpretable GDL models, validated on real scientific datasets, and grounded in the information bottleneck principle.

Findings

LRI models better align detected points with ground-truth scientific patterns.

LRI enhances model robustness to distribution shifts.

LRI provides more stable and meaningful interpretations than post-hoc methods.

Abstract

Point cloud data is ubiquitous in scientific fields. Recently, geometric deep learning (GDL) has been widely applied to solve prediction tasks with such data. However, GDL models are often complicated and hardly interpretable, which poses concerns to scientists who are to deploy these models in scientific analysis and experiments. This work proposes a general mechanism, learnable randomness injection (LRI), which allows building inherently interpretable models based on general GDL backbones. LRI-induced models, once trained, can detect the points in the point cloud data that carry information indicative of the prediction label. We also propose four datasets from real scientific applications that cover the domains of high-energy physics and biochemistry to evaluate the LRI mechanism. Compared with previous post-hoc interpretation methods, the points detected by LRI align much better and stabler with the ground-truth patterns that have actual scientific meanings. LRI is grounded by the information bottleneck principle, and thus LRI-induced models are also more robust to distribution shifts between training and test scenarios. Our code and datasets are available at \url{https://github.com/Graph-COM/LRI}.