Preprint · 2018 · Open access
Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes
arXiv (Cornell University) · Published 11 October 2018 · DOI 10.48550/arxiv.1810.05148
In brief
“Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes” is a 2018 preprint by Novak, Roman and 8 co-authors, published in arXiv (Cornell University). OpenAlex counts 308 citations as of 18 November 2026. It is open access: the full text is free to read at arXiv (Cornell University).
- Citations
- 308
- Most cited in
- 2021 (62)
- References
- 73
- Authors
- 9
Abstract
There is a previously identified equivalence between wide fully connected neural networks (FCNs) and Gaussian processes (GPs). This equivalence enables, for instance, test set predictions that would have resulted from a fully Bayesian, infinitely wide trained FCN to be computed without ever instantiating the FCN, but by instead evaluating the corresponding GP. In this work, we derive an analogous equivalence for multi-layer convolutional neural networks (CNNs) both with and without pooling layers, and achieve state of the art results on CIFAR10 for GPs without trainable kernels. We also introduce a Monte Carlo method to estimate the GP corresponding to a given neural network architecture, even in cases where the analytic form has too many terms to be computationally feasible. Surprisingly, in the absence of pooling layers, the GPs corresponding to CNNs with and without weight sharing are identical. As a consequence, translation equivariance, beneficial in finite channel CNNs trained with stochastic gradient descent (SGD), is guaranteed to play no role in the Bayesian treatment of the infinite channel limit - a qualitative difference between the two regimes that is not present in the FCN case. We confirm experimentally, that while in some scenarios the performance of SGD-trained finite CNNs approaches that of the corresponding GPs as the channel count increases, with careful tuning SGD-trained CNNs can significantly outperform their corresponding GPs, suggesting advantages from SGD training compared to fully Bayesian parameter estimation.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Roman, N., Xiao, L., Lee, J., Bahri, Y., Yang, G., Hron, J., Abolafia, D. A., Pennington, J., & Sohl‐Dickstein, J. (2018). Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes. arXiv (Cornell University). https://doi.org/10.48550/arxiv.1810.05148
MLA 9
Roman, Novak,, et al. “Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes.” arXiv (Cornell University), 2018. https://doi.org/10.48550/arxiv.1810.05148.
Chicago
Roman, Novak,, Lechao Xiao, Jaehoon Lee, Yasaman Bahri, Greg Yang, Jiri Hron, Daniel A. Abolafia, Jeffrey Pennington, and Jascha Sohl‐Dickstein. 2018. “Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.1810.05148.
Harvard
Roman, N. et al. (2018) ‘Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes’, arXiv (Cornell University). doi:10.48550/arxiv.1810.05148.
IEEE
N. Roman et al., “Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes,” arXiv (Cornell University), 2018, doi: 10.48550/arxiv.1810.05148.
BibTeX
@misc{roman2018bayesian,
title = {{Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes}},
author = {Roman, Novak, and Xiao, Lechao and Lee, Jaehoon and Bahri, Yasaman and Yang, Greg and Hron, Jiri and Abolafia, Daniel A. and Pennington, Jeffrey and Sohl‐Dickstein, Jascha},
howpublished = {arXiv (Cornell University)},
year = {2018},
doi = {10.48550/arxiv.1810.05148},
url = {https://doi.org/10.48550/arxiv.1810.05148}
}
Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)
Citations by year
| Year | Citations | Share of the best year |
|---|---|---|
| 2018 | 4 | |
| 2019 | 31 | |
| 2020 | 58 | |
| 2021 | 62 | |
| 2022 | 42 | |
| 2023 | 38 | |
| 2024 | 37 | |
| 2025 | 24 | |
| 2026 | 12 |
Topic and keywords
Physical Sciences › Computer Science › Artificial Intelligence › Gaussian Processes and Bayesian Inference
- infinite-width neural networks
- Gaussian processes
- convolutional neural networks
- pooling layers
- parameter sharing
- translation equivariance
- stochastic gradient descent
- Bayesian neural networks
Work with this paper
Most-cited papers in Computer Science
- ImageNet classification with deep convolutional neural networksAlex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton · Communications of the ACM · 2017 · 114,726 citations
- Long Short-Term MemorySepp Hochreiter, Jürgen Schmidhuber · Neural Computation · 1997 · 109,097 citations
- A Mathematical Theory of CommunicationClaude E. Shannon · Bell System Technical Journal · 1948 · 91,736 citations
- Deep learningYann LeCun, Yoshua Bengio, Geoffrey E. Hinton · Nature · 2015 · 90,226 citations
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksShaoqing Ren, Kaiming He, Ross Girshick and 1 more · IEEE Transactions on Pattern Analysis and Machine Intelligence · 2016 · 74,143 citations
Questions about this paper
How do I cite this paper in APA style?
Roman, N., Xiao, L., Lee, J., Bahri, Y., Yang, G., Hron, J., Abolafia, D. A., Pennington, J., & Sohl‐Dickstein, J. (2018). Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processes. arXiv (Cornell University). https://doi.org/10.48550/arxiv.1810.05148
How many times has this paper been cited?
308 times, according to OpenAlex on 18 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.
Is this paper free to read?
Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.
Who wrote this paper?
Novak, Roman, Lechao Xiao, Jaehoon Lee, Yasaman Bahri, Greg Yang, Jiri Hron, Daniel A. Abolafia, Jeffrey Pennington, Jascha Sohl‐Dickstein.
What is the DOI of this paper?
10.48550/arxiv.1810.05148, permanently at https://doi.org/10.48550/arxiv.1810.05148. Sign in and open it from this page.
Bibliographic data from OpenAlex (CC0), last checked 18 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.