# On the validity of memristor modeling in the neural network literature

## TL;DR

This paper critically examines existing models labeled as memristive in neural network research, revealing that many do not accurately represent true memristive behavior, thus questioning their validity.

## Contribution

It clarifies the distinction between genuine memristive models and non-memristive alternatives used in neural network literature, highlighting the need for proper modeling.

## Key findings

- Many models are non-memristive, describing resistors or bi-state systems.
- A significant portion of literature uses models unrelated to true memristors.
- Questionable relevance of some published results to actual memristive neural networks.

## Abstract

An analysis of the literature shows that there are two types of non-memristive models that have been widely used in the modeling of so-called "memristive" neural networks. Here, we demonstrate that such models have nothing in common with the concept of memristive elements: they describe either non-linear resistors or certain bi-state systems, which all are devices without memory. Therefore, the results presented in a significant number of publications are at least questionable, if not completely irrelevant to the actual field of memristive neural networks.