arXiv：cs.LG（机器学习，全量分类）· Nazmus Saadat As-Saquib, A N M Nafiz Abeer, Hung-Ta Chien, Byung-Jun Yoon, Suhas Kumar, Su-in Yi·· 15 小时前AI 评分31
Forward Target Propagation：一种基于局部损失的前向全局误差信用分配方法
Forward Target Propagation: A Forward-Only Approach to Global Error Credit Assignment via Local Losses
AI 导读
研究者提出 Forward Target Propagation（FTP），用第二次前向传播替代反向传播，仅靠前馈计算估计逐层目标，无需对称反馈权重或可学习逆函数。FTP 在 MNIST、CIFAR10、CIFAR100 上的全连接网络、CNN 和 RNN 中取得与 BP 相当的精度，并在量化低精度硬件约束下优于 BP，效率也高于其他生物启发方法。
正文
Abstract:Training neural networks has traditionally relied on backpropagation (BP), a gradient-based algorithm that, despite its widespread success, suffers from key limitations in both biological and hardware perspectives. These include backward error propagation by symmetric weights, non-local credit assignment, and frozen activity during backward passes. We propose Forward Target Propagation (FTP), a biologically plausible and computationally efficient alternative that replaces the backward pass with a second forward pass. FTP estimates layerwise targets using only feedforward computations, eliminating the need for symmetric feedback weights or learnable inverse functions, hence enabling modular and local learning. We evaluate FTP on fully connected networks, CNNs, and RNNs, demonstrating accuracies competitive with BP on MNIST, CIFAR10, and CIFAR100, as well as effective modeling of long-term dependencies in sequential tasks. Moreover, FTP outperforms BP under quantized low-precision and emerging hardware constraints while also demonstrating substantial efficiency gains over other biologically inspired methods such as target propagation variants and forward-only learning algorithms. With its minimal computational overhead, forward-only nature, and hardware compatibility, FTP provides a promising direction for energy-efficient on-device learning and neuromorphic computing.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2506.11030 [cs.LG] |
| (or arXiv:2506.11030v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2506.11030 arXiv-issued DOI via DataCite |
Submission history
From: Nazmus Saadat As -Saquib [view email]
[v1]
Tue, 20 May 2025 16:09:23 UTC (6,962 KB)
[v2]
Thu, 1 Oct 2026 15:11:54 UTC (6,853 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org