Center for Theoretical Neuroscience, Columbia University, USA; Department of Brain and Cognitive Sciences, MIT, USA; Department of Electrical Engineering and Computer Science, MIT, USA.
Brain Circuits & Behavior Lab, IDIBAPS, Barcelona, Spain.
Curr Opin Neurobiol. 2021 Oct;70:182-192. doi: 10.1016/j.conb.2021.10.015. Epub 2021 Nov 26.
Recurrent neural networks (RNNs) trained with machine learning techniques on cognitive tasks have become a widely accepted tool for neuroscientists. In this short opinion piece, we discuss fundamental challenges faced by the early work of this approach and recent steps to overcome such challenges and build next-generation RNN models for cognition. We propose several essential questions that practitioners of this approach should address to continue to build future generations of RNN models.
基于机器学习技术在认知任务上训练的递归神经网络 (RNN) 已经成为神经科学家广泛接受的工具。在这篇简短的观点文章中，我们讨论了该方法早期工作所面临的基本挑战，以及最近克服这些挑战并构建新一代用于认知的 RNN 模型的步骤。我们提出了该方法的从业者应该解决的几个基本问题，以继续构建新一代 RNN 模型。
