# A Biologically Plausible Learning Rule for Deep Learning in the Brain

arXiv: https://arxiv.org/abs/1811.01768 · HTML: https://mlsanity.com/paper/1811.01768/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Nov 2018.

## Built on by (most cited first, 9 of 9)

- [Brain-Inspired Machine Intelligence: A Survey of Neurobiologically-Plausible Credit Assignment](https://mlsanity.com/paper/2312.09257/) (2023)
- [Learning by non-interfering feedback chemical signaling in physical networks](https://mlsanity.com/paper/2203.12098/) (2022)
- [Learning efficient backprojections across cortical hierarchies in real time](https://mlsanity.com/paper/2212.10249/) (2022)
- [Spiking Neural Networks and Bio-Inspired Supervised Deep Learning: A Survey](https://mlsanity.com/paper/2307.16235/) (2023)
- [Fast weight programming and linear transformers: from machine learning to neurobiology](https://mlsanity.com/paper/2508.08435/) (2025)
- [Beyond accuracy: generalization properties of bio-plausible temporal credit assignment rules](https://mlsanity.com/paper/2206.00823/) (2022)
- [Of Mice and Machines: A Comparison of Learning Between Real World Mice and RL Agents](https://mlsanity.com/paper/2505.12204/) (2025)
- [Spike-based alignment learning solves the weight transport problem](https://mlsanity.com/paper/2503.02642/) (2025)
- [An energy-efficient spiking neural network with continuous learning for self-adaptive brain-machine interface](https://mlsanity.com/paper/2511.22108/) (2025)

## Abstract (arXiv)

Researchers have proposed that deep learning, which is providing important progress in a wide range of high complexity tasks, might inspire new insights into learning in the brain. However, the methods used for deep learning by artificial neural networks are biologically unrealistic and would need to be replaced by biologically realistic counterparts. Previous biologically plausible reinforcement learning rules, like AGREL and AuGMEnT, showed promising results but focused on shallow networks with three layers. Will these learning rules also generalize to networks with more layers and can they handle tasks of higher complexity? We demonstrate the learning scheme on classical and hard image-classification benchmarks, namely MNIST, CIFAR10 and CIFAR100, cast as direct reward tasks, both for fully connected, convolutional and locally connected architectures. We show that our learning rule - Q-AGREL - performs comparably to supervised learning via error-backpropagation, with this type of trial-and-error reinforcement learning requiring only 1.5-2.5 times more epochs, even when classifying 100 different classes as in CIFAR100. Our results provide new insights into how deep learning may be implemented in the brain.