Learning depends on changes in synaptic connections deep inside the brain. In multilayer networks, these changes are triggered by error signals fed back from the output, generally through a stepwise inversion of the feedforward processing steps. The gold standard for this process -- backpropagation -- works well in artificial neural networks, but is biologically implausible. Several recent proposals have emerged to address this problem, but many of these biologically-plausible schemes are based on learning an independent set of feedback connections. This complicates the assignment of errors to each synapse by making it dependent upon a second learning problem, and by fitting inversions rather than guaranteeing them. Here, we show that feedforward network transformations can be effectively inverted through dynamics. We derive this dynamic inversion from the perspective of feedback control, where the forward transformation is reused and dynamically interacts with fixed or random feedback to propagate error signals during the backward pass. Importantly, this scheme does not rely upon a second learning problem for feedback because accurate inversion is guaranteed through the network dynamics. We map these dynamics onto generic feedforward networks, and show that the resulting algorithm performs well on several supervised and unsupervised datasets. We also link this dynamic inversion to Gauss-Newton optimization, suggesting a biologically-plausible approximation to second-order learning. Overall, our work introduces an alternative perspective on credit assignment in the brain, and proposes a special role for temporal dynamics and feedback control during learning.

中文翻译：

---

通过前馈网络的动态反演进行生物信用分配

学习取决于大脑深处突触连接的变化。在多层网络中，这些变化是由输出反馈的错误信号触发的，通常是通过前馈处理步骤的逐步反转。这个过程的黄金标准——反向传播——在人工神经网络中运行良好，但在生物学上是不可信的。最近出现了一些解决这个问题的提议，但许多这些生物学上合理的方案都是基于学习一组独立的反馈连接。这使每个突触的错误分配变得复杂，因为它依赖于第二个学习问题，并且通过拟合反转而不是保证它们。在这里，我们表明前馈网络转换可以通过动力学有效地反转。我们从反馈控制的角度推导出这种动态反转，其中前向变换被重用，并与固定或随机反馈动态交互，以在后向传递期间传播误差信号。重要的是，该方案不依赖于第二个学习问题的反馈，因为通过网络动态保证了准确的反演。我们将这些动态映射到通用前馈网络上，并表明所得算法在几个监督和非监督数据集上表现良好。我们还将这种动态反演与 Gauss-Newton 优化联系起来，这表明二阶学习在生物学上是合理的。总的来说，我们的工作引入了大脑中学分分配的另一种视角，并提出了学习过程中时间动态和反馈控制的特殊作用。