Towards Global Neural Network Abstractions with Locally-Exact Reconstruction

TL;DR

This paper introduces GINNACER, a novel neural network abstraction method that provides tight, global over-approximation bounds with exact local reconstructions, improving over existing techniques in safety and explainability.

Contribution

GINNACER is the first approach to achieve both global tight bounds and local exact reconstructions in neural network abstraction.

Findings

GINNACER outperforms existing global abstraction methods in tightness.

GINNACER is competitive with local abstraction techniques.

Experiments demonstrate GINNACER's scalability and accuracy.

Abstract

Neural networks are a powerful class of non-linear functions. However, their black-box nature makes it difficult to explain their behaviour and certify their safety. Abstraction techniques address this challenge by transforming the neural network into a simpler, over-approximated function. Unfortunately, existing abstraction techniques are slack, which limits their applicability to small local regions of the input domain. In this paper, we propose Global Interval Neural Network Abstractions with Center-Exact Reconstruction (GINNACER). Our novel abstraction technique produces sound over-approximation bounds over the whole input domain while guaranteeing exact reconstructions for any given local input. Our experiments show that GINNACER is several orders of magnitude tighter than state-of-the-art global abstraction techniques, while being competitive with local ones.