This paper demonstrates how neural networks can be applied to model and predict the functional behaviour of disordered nano-particle and nano-tube networks. In recently published experimental work, nano-particle and nano-tube networks show promising functionality for future reconfigurable devices, without a predefined design. The nano-material has been treated as a black-box, and the principle of evolution-in-materio, involving genetic algorithms, has been used to find appropriate configuration voltages to enable the target functionality. In order to support future experiments and the development of useful devices based on disordered nano-materials, we developed simulation tools for predicting candidate functionalities. One of these tools is based on a physical model, but the one described and analysed in this paper is based on an artificial neural network model. The advantage of this newly presented approach is that, after training the neural network to match either the real material or its physical model, it can be configured using gradient descent instead of a black-box optimisation, speeding up the search for functionality. The neural networks do not simulate the physical properties, but rather approximate the nano material’s transfer functions. The functions found using this new technique were verified back on the nano material’s physical model and on a real material network. It can be concluded from the reported experiments with these neural network models that they model the simulated nano-material quite accurately. The differentiable, neural network-based material model is used to find logic gates, as a proof of principle. This shows that the new approach has great potential for partly replacing costly and time consuming experiments with the real nano-material. Therefore, this approach has a high relevance for future computing, either as an alternative to digital computing or as an alternative way of producing multifunctional reconfigurable devices.