Abstract | In the artificial intelligence subfield of neural networks , a barrier to that goal is that when agents learn a new skill they typically do so by losing previously acquired skills, a problem called catastrophic forgetting. |
Abstract | In this paper, we test whether catastrophic forgetting can be reduced by evolving modular neural networks . |
Abstract | To produce modularity, we evolve neural networks with a cost for neural connections. |
Author Summary | A longstanding goal in artificial intelligence (AI) is creating computational brain models ( neural networks ) that learn what to do in new situations. |
Author Summary | Here we test whether such forgetting is reduced by evolving modular neural networks , meaning networks with many distinct subgroups of neurons. |
Background | Catastrophic forgetting (also called catastrophic interference) has been identified as a problem for artificial neural networks (ANNs) for over two decades: When learning multiple tasks in a sequence, previous skills are forgotten rapidly as new information is learned [9, 10]. |
Introduction | Such forgetting is especially problematic in fields that attempt to create artificial intelligence in brain models called artificial neural networks [1, 4, 5]. |
Introduction | To learn new skills, neural network learning algorithms change the weights of neural connections [6—8], but old skills are lost because the weights that encoded old skills are changed to improve performance on new tasks. |
Introduction | To advance our goal of producing sophisticated, functional artificial intelligence in neural networks and make progress in our longterm quest to create general artificial intelligence with them, we need to develop algorithms that can learn how to handle more than a few different problems. |
AMSN | This scenario is shown for the firing rate model in Fig 1B and for the spiking neural network in Fig 4). |
Effect of cortical spiking activity correlations on the DTT | These simulations were only performed for the spiking neural network model since modelling correlations in a mean field model is nontrivial, especially when post-synaptic neurons are recurrently connected. |
Supporting Information | Presence of DT T in spiking neural network in which D1 and D2 MSNs have different F-I curves. |