In this study, we aim to interpret the hidden semantics of each unit within convolutional neural networks by abstracting local activated patterns of each neuron into corresponding global functions. Unlike existing quantitative hidden-semantics-based explanations which require comparing pixel-wise annotated data one by one, the proposed active interpretability method gives function embeddings after the training without semantic annotation. Specifically, the function of a neuron is denoted as its global expected activated pattern, and therefore feature space and function embedding space are unsupervised aligned during training. The synchronization mechanism is introduced to aggregate scattered function embeddings into function modules, transforming the excessive gray-box interpretations into white-box ones. Moreover, the hard routing guided by function embedding is employed to ensure semantic specificity. We explore the aggregated function modules to showcase the qualitative interpretability of functionally motivated networks. Meanwhile, the proposed method exhibits superior quantitative interpretability metrics such as accuracy, faithfulness, robustness, and complexity.

Synchronization is a ubiquitous phenomenon in nature that enables the orderly presentation of information. In the human brain, for instance, functional modules such as the visual, motor, and language cortices form through neuronal synchronization. Inspired by biological brains and previous neuroscience studies, we propose an interpretable neural network incorporating a synchronization mechanism. The basic idea is to constrain each neuron, such as a convolution filter, to capture a single semantic pattern while synchronizing similar neurons to facilitate the formation of interpretable functional modules. Specifically, we regularize the activation map of a neuron to surround its focus position of the activated pattern in a sample. Moreover, neurons locally interact with each other, and similar ones are synchronized together during the training phase adaptively. Such local aggregation preserves the globally distributed representation nature of the neural network model, enabling a reasonably interpretable representation. To analyze the neuron interpretability comprehensively, we introduce a series of novel evaluation metrics from multiple aspects. Qualitative and quantitative experiments demonstrate that the proposed method outperforms many state-of-the-art algorithms in terms of interpretability. The resulting synchronized functional modules show module consistency across data and semantic specificity within modules.

In this paper, we introduce a novel method to discover common and distinct structural connectivity patterns between SZP and MDD via a Cluster-Driven Nonnegative Matrix Factorization (called CD-NMF). Specifically, CD-NMF is applied to decompose the joint structural connectivity map into common and distinct parts, and each part is further factorized into two sub-matrices (i.e. common/distinct basis matrix and common/distinct encoding matrix) correspondingly. By imposing the clustering constraints on common and distinct encoding matrices, the discriminative patterns as well as the common patterns between the two disorders are extracted simultaneously. Experimental results demonstrate that CD-NMF allows finding the common and distinct structural patterns effectively. More importantly, the derived distinct patterns, show powerful ability to discriminate the patients of schizophrenia and major depressive disorder.