Purin：一种面向人工神经网络的仿生机制
Purin: A Biology-inspired Mechanism for Artificial Neural Networks
AI 导读
研究者提出 Purin，一种仿生且兼容常规卷积神经网络的突触效能调制机制，通过基于时间间隔的抽象引入短期与长期突触效能变化，无需离散时间步。该机制以有界因子和输入、输出两侧两个权重矩阵表示效能变化，权重经反向传播更新。实验显示，在去除混淆因素后的 AlexNet、VGG11 和 GoogLeNet 上，Purin 在所有评测数据集上都提升了分类准确率。
正文 · 原文
Abstract:Artificial neural networks (ANNs) usually represent neural transmission with fixed trainable weights during a training batch, which omits short-term changes in synaptic efficacy. In addition, the discrete time-step simulation requires additional temporal processing that many conventional ANN architectures do not use. To overcome these challenges, we propose Purin, a biology-inspired and ANN-compatible mechanism, that introduces synaptic efficacy modulation into conventional convolutional neural networks. Purin uses a time-interval-based abstraction for neural activities, which allows Purin to introduce short- and long-term synaptic efficacy changes without using discrete time-steps. Purin introduces a bounded factor to represent temporary synaptic efficacy changes, together with two weight matrices that represent input-side and output-side efficacy. The weight matrices are updated by backpropagation and interpreted as the long-term synaptic efficacy changes. Experimental results show that after removing the confounding factors in the AlexNet, VGG11, and GoogLeNet architectures, Purin improves the classification accuracies in all three models across the evaluated datasets.
| Comments: | 8 pages, 2 figures, 8 tables |
| Subjects: | Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE) |
| MSC classes: | 68T05 |
| ACM classes: | I.2.6 |
| Cite as: | arXiv:2609.31235 [cs.AI] |
| (or arXiv:2609.31235v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2609.31235 arXiv-issued DOI via DataCite |
Submission history
From: Zishu Liu [view email]
[v1]
Fri, 25 Sep 2026 13:15:22 UTC (919 KB)
[v2]
Mon, 9 Nov 2026 21:21:07 UTC (1,307 KB)
来源：arXiv cs.AI · arxiv.org