Methods › Computer Vision › Convolutional Neural Networks › RevNet
RevNet
Introduced by Aidan N. Gomez et al. in The Reversible Residual Network: Backpropagation Without Storing Activations
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
A Reversible Residual Network, or RevNet, is a variant of a ResNet where each layer’s activations can be reconstructed exactly from the next layer’s. Therefore, the activations for most layers need not be stored in memory during backpropagation. The result is a network architecture whose activation storage requirements are independent of depth, and typically at least an order of magnitude smaller compared with equally sized ResNets.
RevNets are composed of a series of reversible blocks. Units in each layer are partitioned into two groups, denoted x₁ and x₂; the authors find what works best is partitioning the channels. Each reversible block takes inputs (x₁, x₂) and produces outputs (y₁, y₂) according to the following additive coupling rules – inspired the transformation in NICE (nonlinear independent components estimation) – and residual functions F and G analogous to those in standard ResNets:
y₁ = x₁ + F(x₂) y₂ = x₂ + G(y₁)
Each layer’s activations can be reconstructed from the next layer’s activations as follows:
x₂ = y₂ − G(y₁) x₁ = y₁ − F(x₂)
Note that unlike residual blocks, reversible blocks must have a stride of 1 because otherwise the layer discards information, and therefore cannot be reversible. Standard ResNet architectures typically have a handful of layers with a larger stride. If we define a RevNet architecture analogously, the activations must be stored explicitly for all non-reversible layers.
Papers archive 2025-07-28
9 shown of 9, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.
-
Diffusion Models Beat GANs on Image Classification 17 Jul 2023 · 1 repository · arXiv:2307.08702
-
Conditional Injective Flows for Bayesian Imaging 15 Apr 2022 · 1 repository · arXiv:2204.07664
-
Level set learning with pseudo-reversible neural networks for nonlinear dimension reduction in function approximation 2 Dec 2021 · 2 repositories · arXiv:2112.01438
-
Multi-split Reversible Transformers Can Enhance Neural Machine Translation 1 Apr 2021 · 0 repositories
-
Object Segmentation Without Labels with Large-Scale Generative Models 8 Jun 2020 · 1 repository · arXiv:2006.04988
-
Reconstructing Natural Scenes from fMRI Patterns using BigBiGAN 31 Jan 2020 · 0 repositories · arXiv:2001.11761
-
Large Scale Adversarial Representation Learning 4 Jul 2019 · 4 repositories · arXiv:1907.02544Syntology ran 0 of 3 samples · 3 unverified
-
Beyond Finite Layer Neural Networks: Bridging Deep Architectures and Numerical Differential Equations 27 Oct 2017 · 0 repositories · arXiv:1710.10121
-
The Reversible Residual Network: Backpropagation Without Storing Activations 14 Jul 2017 · 9 repositories · arXiv:1707.04585Syntology ran 4 of 4 samples · 0 unverified
Tasks archive 2025-07-28
20 shown of 32 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.
Usage over time archive 2025-07-28
Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).
Categories archive 2025-07-28
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections