# On the relationship between predictive coding and backpropagation

Article · Robert Rosenbaum

- Archive id: arxiv:2106.13082
- Published: 2021-06-20
- Venue: PLoS ONE
- DOI: https://doi.org/10.1371/journal.pone.0266102
- arXiv: https://arxiv.org/abs/2106.13082
- Citations: 20
- Source URL: https://arxiv.org/abs/2106.13082
- PDF: https://arxiv.org/pdf/2106.13082
- Provenance: arxiv

## Abstract

Artificial neural networks are often interpreted as abstract models of biological neuronal networks, but they are typically trained using the biologically unrealistic backpropagation algorithm and its variants. Predictive coding has been proposed as a potentially more biologically realistic alternative to backpropagation for training neural networks. This manuscript reviews and extends recent work on the mathematical relationship between predictive coding and backpropagation for training feedforward artificial neural networks on supervised learning tasks. Implications of these results for the interpretation of predictive coding and deep neural networks as models of biological learning are discussed along with a repository of functions, Torch2PC, for performing predictive coding with PyTorch neural network models.

_Abstract reproduced under CC0-1.0 (arXiv descriptive metadata)._

Topics: [NeuroAI & Computational Neuroscience](https://space-kitty.com/topics/neuroai)

Cited in dossiers:
- [NeuroAI & Computational Neuroscience: State of the Field (October 2026)](https://space-kitty.com/dossiers/neuroai)

---
Space Kitty Archive · https://space-kitty.com · API: https://space-kitty.com/api/v1 · MCP: https://space-kitty.com/mcp · Premium: $0.20/request via x402 (Base Sepolia (testnet)). Metadata: arXiv (CC0), OpenAlex (CC0), Europe PMC (CC BY/CC0 works), ClinicalTrials.gov.