Understanding Memories of the Past in the Context of Different Complex Neural Network Architectures

Self Cite

Deep learning models have achieved an impressive performance in a variety of tasks, but they often suffer from overfitting and are vulnerable to adversarial attacks. Previous research has shown that dropout regularization is an effective technique that can improve model generalization and robustness. In this study, we investigate the impact of dropout regularization on the ability of neural networks to withstand adversarial attacks, as well as the degree of “functional smearing” between individual neurons in the network. Functional smearing in this context describes the phenomenon that a neuron or hidden state is involved in multiple functions at the same time. Our findings confirm that dropout regularization can enhance a network’s resistance to adversarial attacks, and this effect is only observable within a specific range of dropout probabilities. Furthermore, our study reveals that dropout regularization significantly increases the distribution of functional smearing across a wide range of dropout rates. However, it is the fraction of networks with lower levels of functional smearing that exhibit greater resilience against adversarial attacks. This suggests that, even though dropout improves robustness to fooling, one should instead try to decrease functional smearing.

Section: Discussionmentioning

confidence: 85%

Section: Introductionmentioning

confidence: 99%

Robustness of Sparsely Distributed Representations to Adversarial Attacks in Deep Neural Networks

Sardar,

Khan,

Hintze

et al. 2023

Self Cite

“…In particular, we are coarse-graining those entropies by mapping all continuous values to the binary categories 0 and 1. We previously used the median value of a neuron’s excitation level as the threshold for the bin [ 45 ]. Instead, here the hidden-state values are clustered using a k -means clustering algorithm with

.…”

Section: Methodsmentioning

confidence: 99%

Detecting Information Relays in Deep Neural Networks

Hintze

Adami

2023

Self Cite

Deep learning of artificial neural networks (ANNs) is creating highly functional processes that are, unfortunately, nearly as hard to interpret as their biological counterparts. Identification of functional modules in natural brains plays an important role in cognitive and neuroscience alike, and can be carried out using a wide range of technologies such as fMRI, EEG/ERP, MEG, or calcium imaging. However, we do not have such robust methods at our disposal when it comes to understanding functional modules in artificial neural networks. Ideally, understanding which parts of an artificial neural network perform what function might help us to address a number of vexing problems in ANN research, such as catastrophic forgetting and overfitting. Furthermore, revealing a network’s modularity could improve our trust in them by making these black boxes more transparent. Here, we introduce a new information-theoretic concept that proves useful in understanding and analyzing a network’s functional modularity: the relay information IR. The relay information measures how much information groups of neurons that participate in a particular function (modules) relay from inputs to outputs. Combined with a greedy search algorithm, relay information can be used to identify computational modules in neural networks. We also show that the functionality of modules correlates with the amount of relay information they carry.

“…We consider two different tasks. The first is a memorization task: the object is to read back a set of inputs with different delays (a task we call "n-Back" [16]). The second task is an active perception task (as pioneered by Beer [17,18]), where the agent has to make a decision on whether to catch or avoid a falling block based on its size and/or direction of motion.…”

Section: Information Fragmentationmentioning

confidence: 99%

“…The first task is n-Back [16], a simple memory problem where an agent is provided a sequence of binary digits through a single sensory node, one at a time. The inputs must be stored in memory and read out with particular delays.…”

Section: N-back Taskmentioning

confidence: 99%

Information Fragmentation, Encryption and Information Flow in Complex Biological Networks

Bohm

Kirkpatrick

Cao

et al. 2022

Self Cite

Assessing where and how information is stored in biological networks (such as neuronal and genetic networks) is a central task both in neuroscience and in molecular genetics, but most available tools focus on the network’s structure as opposed to its function. Here, we introduce a new information-theoretic tool—information fragmentation analysis—that, given full phenotypic data, allows us to localize information in complex networks, determine how fragmented (across multiple nodes of the network) the information is, and assess the level of encryption of that information. Using information fragmentation matrices we can also create information flow graphs that illustrate how information propagates through these networks. We illustrate the use of this tool by analyzing how artificial brains that evolved in silico solve particular tasks, and show how information fragmentation analysis provides deeper insights into how these brains process information and “think”. The measures of information fragmentation and encryption that result from our methods also quantify complexity of information processing in these networks and how this processing complexity differs between primary exposure to sensory data (early in the lifetime) and later routine processing.