# aSTDP: A More Biologically Plausible Learning

arXiv: https://arxiv.org/abs/2206.14137 · HTML: https://mlsanity.com/paper/2206.14137/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 24 references that are in the map.
- First posted on arXiv 22 May 2022.

## Builds on (most cited first, 8 of 8)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [BEiT: BERT Pre-Training of Image Transformers](https://mlsanity.com/paper/2106.08254/) (2021)
- [Difference Target Propagation](https://mlsanity.com/paper/1412.7525/) (2014)
- [Towards Biologically Plausible Deep Learning](https://mlsanity.com/paper/1502.04156/) (2015)
- [Random feedback weights support learning in deep neural networks](https://mlsanity.com/paper/1411.0247/) (2014)
- [Early Inference in Energy-Based Models Approximates Back-Propagation](https://mlsanity.com/paper/1510.02777/) (2015)
- [STDP as presynaptic activity times rate of change of postsynaptic activity](https://mlsanity.com/paper/1509.05936/) (2015)
- [Contrastive Hebbian Learning with Random Feedback Weights](https://mlsanity.com/paper/1806.07406/) (2018)

## Abstract (arXiv)

Spike-timing dependent plasticity in biological neural networks has been proven to be important during biological learning process. On the other hand, artificial neural networks use a different way to learn, such as Back-Propagation or Contrastive Hebbian Learning. In this work we introduce approximate STDP, a new neural networks learning framework more similar to the biological learning process. It uses only STDP rules for supervised and unsupervised learning, every neuron distributed learn patterns and don' t need a global loss or other supervised information. We also use a numerical way to approximate the derivatives of each neuron in order to better use SDTP learning and use the derivatives to set a target for neurons to accelerate training and testing process. The framework can make predictions or generate patterns in one model without additional configuration. Finally, we verified our framework on MNIST dataset for classification and generation tasks.