NeuroView: Explainable Deep Network Decision Making
arXiv:2110.07778
Abstract
Deep neural networks (DNs) provide superhuman performance in numerous computer vision tasks, yet it remains unclear exactly which of a DN's units contribute to a particular decision. NeuroView is a new family of DN architectures that are interpretable/explainable by design. Each member of the family is derived from a standard DN architecture by vector quantizing the unit output values and feeding them into a global linear classifier. The resulting architecture establishes a direct, causal link between the state of each unit and the classification decision. We validate NeuroView on standard datasets and classification tasks to show that how its unit/class mapping aids in understanding the decision-making process.
arXiv comment12 pages, 7 figures
References in corpus (16)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Axiomatic Attribution for Deep Networks
- Learning Important Features Through Propagating Activation Differences
- Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1
- Model-Agnostic Interpretability of Machine Learning
- Transfusion: Understanding Transfer Learning for Medical Imaging
- Understanding the Role of Individual Units in a Deep Neural Network
- Multi-view Convolutional Neural Networks for 3D Shape Recognition
- Functional linear regression that's interpretable
- GAN Dissection: Visualizing and Understanding Generative Adversarial Networks
- Do CIFAR-10 Classifiers Generalize to CIFAR-10?
- Rethinking CNN Models for Audio Classification
- Hierarchical interpretations for neural network predictions
- Explaining Classifiers with Causal Concept Effect (CaCE)
- Neuron Shapley: Discovering the Responsible Neurons
- Learning from Failure: Training Debiased Classifier from Biased Classifier