DeepAI AI Chat
Log In Sign Up

Higher-order mutual information reveals synergistic sub-networks for multi-neuron importance

by   Kenzo Clauw, et al.

Quantifying which neurons are important with respect to the classification decision of a trained neural network is essential for understanding their inner workings. Previous work primarily attributed importance to individual neurons. In this work, we study which groups of neurons contain synergistic or redundant information using a multivariate mutual information method called the O-information. We observe the first layer is dominated by redundancy suggesting general shared features (i.e. detecting edges) while the last layer is dominated by synergy indicating local class-specific features (i.e. concepts). Finally, we show the O-information can be used for multi-neuron importance. This can be demonstrated by re-training a synergistic sub-network, which results in a minimal change in performance. These results suggest our method can be used for pruning and unsupervised representation learning.


page 1

page 2

page 3

page 4


Understanding Individual Neuron Importance Using Information Theory

In this work, we characterize the outputs of individual neurons in a tra...

NISP: Pruning Networks using Neuron Importance Score Propagation

To reduce the significant redundancy in deep Convolutional Neural Networ...

Natural Language Descriptions of Deep Visual Features

Some neurons in deep networks specialize in recognizing highly specific ...

Quantifying High-order Interdependencies via Multivariate Extensions of the Mutual Information

This article introduces a model-agnostic approach to study statistical s...

Understanding trained CNNs by indexing neuron selectivity

The impressive performance and plasticity of convolutional neural networ...

Invariant Representations with Stochastically Quantized Neural Networks

Representation learning algorithms offer the opportunity to learn invari...

Computational analysis of laminar structure of the human cortex based on local neuron features

In this paper, we present a novel method for analysis and segmentation o...