# Learning in the Machine: To Share or Not to Share?

arXiv: https://arxiv.org/abs/1909.11483 · HTML: https://mlsanity.com/paper/1909.11483/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Sep 2019.

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

- [A Rainbow in Deep Network Black Boxes](https://mlsanity.com/paper/2305.18512/) (2023)
- [Towards Biologically Plausible Convolutional Networks](https://mlsanity.com/paper/2106.13031/) (2021)
- [Understanding MLP-Mixer as a Wide and Sparse MLP](https://mlsanity.com/paper/2306.01470/) (2023)
- [Tourbillon: a Physically Plausible Neural Architecture](https://mlsanity.com/paper/2107.06424/) (2021)
- [A Spatially Masked Adaptive Gated Network for multimodal post-flood water extent mapping using SAR and incomplete multispectral data](https://mlsanity.com/paper/2601.00123/) (2025)
- [Detecting Pulmonary Coccidioidomycosis (Valley fever) with Deep Convolutional Neural Networks](https://mlsanity.com/paper/2102.00280/) (2021)
- [The Role Of Biology In Deep Learning](https://mlsanity.com/paper/2209.04425/) (2022)
- [A Theory of Synaptic Neural Balance: From Local to Global Order](https://mlsanity.com/paper/2405.09688/) (2024)
- [Integration of TinyML and LargeML: A Survey of 6G and Beyond](https://mlsanity.com/paper/2505.15854/) (2025)

## Abstract (arXiv)

Weight-sharing is one of the pillars behind Convolutional Neural Networks and their successes. However, in physical neural systems such as the brain, weight-sharing is implausible. This discrepancy raises the fundamental question of whether weight-sharing is necessary. If so, to which degree of precision? If not, what are the alternatives? The goal of this study is to investigate these questions, primarily through simulations where the weight-sharing assumption is relaxed. Taking inspiration from neural circuitry, we explore the use of Free Convolutional Networks and neurons with variable connection patterns. Using Free Convolutional Networks, we show that while weight-sharing is a pragmatic optimization approach, it is not a necessity in computer vision applications. Furthermore, Free Convolutional Networks match the performance observed in standard architectures when trained using properly translated data (akin to video). Under the assumption of translationally augmented data, Free Convolutional Networks learn translationally invariant representations that yield an approximate form of weight sharing.