arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2406.14723cs.NE
收藏

使用预测编码的Hopfield网络在线训练

Online Training of Hopfield Networks using Predictive Coding

Ehsan Ganjidoost, Mallory Snow, Jeff Orchard

AI总结:

本研究首次将预测编码(PC)学习规则直接应用于Hopfield网络(一种循环神经网络)的训练,无需修改规则也无需时间展开,验证了训练后网络表现与经典Hopfield网络一致,拓展了生物可解释学习规则的应用场景。

摘要:

神经科学与人工智能(AI)一直协同发展,二者既增进了我们对大脑的理解,也为生物可解释神经网络(NNs)及学习规则的最新研究提供了启发。预测编码(PC)及其学习规则已被证明能以符合生物学机制的方式近似误差反向传播,其权重更新仅依赖突触前和突触后神经元的活动,具备局部性。与信息单向流动的传统前馈神经网络不同,PC模型通过双向传递信息更精准地模拟大脑:一个方向传输预测信号,另一方向传输校正/误差信号。PC模型的学习方式是将部分神经元钳位到目标值,再运行网络直至达到平衡状态。在平衡状态下,网络会在权重更新所需的位置直接计算自身的误差梯度。传统反向传播要求计算图为前馈结构,但PC版本的反向传播无此限制。令人意外的是,目前尚无研究证明PC学习可直接应用于循环神经网络(RNNs)。Hopfield网络(HNs)是一类实现内容寻址存储器的循环神经网络,能够学习模式(或称“记忆”),并可从部分或受损的模式中检索出完整内容。本文证明,Hopfield网络可直接使用PC学习规则进行训练,无需对规则做任何修改。据我们所知,这是首次将PC学习直接应用于循环神经网络训练,且无需在时间维度上对网络进行展开。实验结果表明,经PC训练的Hopfield网络表现与经典Hopfield网络一致。

英文摘要:

Neuroscience and Artificial Intelligence (AI) have progressed in tandem, each contributing to our understanding of the brain, and inspiring recent developments in biologically-plausible neural networks (NNs) and learning rules. Predictive coding (PC), and its learning rule, have been shown to approximate error backpropagation in a biologically relevant manner, with local weight updates that depend only on the activity of the pre- and post-synaptic neurons. Unlike traditional feedforward NNs where the flow of information goes in one direction, PC models mimic the brain more accurately by passing information bidirectionally: prediction in one direction, and correction/error in the other. PC models learn by clamping some neurons to target values and running the network to equilibrium. At equilibrium, the network calculates its own error gradients right at the location where they are used for weight updates. Traditional backprop requires the computation graph to be feedforward. However, the PC version of backprop does not have this requirement. Amazingly, no one has demonstrated the application of PC learning directly to recurrent neural networks (RNNs). Hopfield networks (HNs) are RNNs that implement a content-addressable memory, learning patterns (or ``memories'') that can be retrieved from partial or corrupted patterns. In this paper, we show that a HN can be trained using the PC learning rules without modification. To our knowledge, this is the first time PC learning has been applied directly to train a RNN, without the need to unroll it in time. Our results indicate that the PC-trained HNs behave like classical HNs.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑