# Decomposer Networks: Deep Component Analysis and Synthesis

arXiv: https://arxiv.org/abs/2510.09825 · HTML: https://mlsanity.com/paper/2510.09825/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 4 of its 15 references that are in the map.
- First posted on arXiv 10 Oct 2025.

## Builds on (most cited first, 4 of 4)

- [A Style-Based Generator Architecture for Generative Adversarial Networks](https://mlsanity.com/paper/1812.04948/) (2018)
- [Object-Centric Learning with Slot Attention](https://mlsanity.com/paper/2006.15055/) (2020)
- [Disentangling by Factorising](https://mlsanity.com/paper/1802.05983/) (2018)
- [Sharing Residual Units Through Collective Tensor Factorization in Deep Neural Networks](https://mlsanity.com/paper/1703.02180/) (2017)

## Abstract (arXiv)

We propose the Decomposer Networks (DecompNet), a semantic autoencoder that factorizes an input into multiple interpretable components. Unlike classical autoencoders that compress an input into a single latent representation, the Decomposer Network maintains N parallel branches, each assigned a residual input defined as the original signal minus the reconstructions of all other branches. By unrolling a Gauss--Seidel style block-coordinate descent into a differentiable network, DecompNet enforce explicit competition among components, yielding parsimonious, semantically meaningful representations. We situate our model relative to linear decomposition methods (PCA, NMF), deep unrolled optimization, and object-centric architectures (MONet, IODINE, Slot Attention), and highlight its novelty as the first semantic autoencoder to implement an all-but-one residual update rule.