论文标题
BORT：朝着具有有限正交约束的可解释神经网络
Bort: Towards Explainable Neural Networks with Bounded Orthogonal Constraint
论文作者
论文摘要
深度学习彻底改变了人类社会，但是深度神经网络的黑盒本质阻碍了可靠性的行业的进一步应用。为了解开它们，许多作品都观察或影响内部变量，以提高黑盒模型的可理解性和可逆性。但是，现有方法依赖于直觉的假设和缺乏数学保证。为了弥合这一差距，我们介绍了Bort，这是一种优化器，用于改善模型性和正交性约束模型参数，从模型可理解性和可逆性的充分条件中得出。我们对通过Bort优化的模型表示进行重建和回溯，并观察到模型解释性的明显改善。基于Bort，我们能够合成可解释的对抗样本，而无需其他参数和培训。令人惊讶的是，我们发现Bort不断提高各种体系结构的分类精度，包括Resnet和Mnist，Cifar-10和Imagenet上的DEIT。代码：https：//github.com/zbr17/bort。
Deep learning has revolutionized human society, yet the black-box nature of deep neural networks hinders further application to reliability-demanded industries. In the attempt to unpack them, many works observe or impact internal variables to improve the comprehensibility and invertibility of the black-box models. However, existing methods rely on intuitive assumptions and lack mathematical guarantees. To bridge this gap, we introduce Bort, an optimizer for improving model explainability with boundedness and orthogonality constraints on model parameters, derived from the sufficient conditions of model comprehensibility and invertibility. We perform reconstruction and backtracking on the model representations optimized by Bort and observe a clear improvement in model explainability. Based on Bort, we are able to synthesize explainable adversarial samples without additional parameters and training. Surprisingly, we find Bort constantly improves the classification accuracy of various architectures including ResNet and DeiT on MNIST, CIFAR-10, and ImageNet. Code: https://github.com/zbr17/Bort.
