# C-Disentanglement: Discovering Causally-Independent Generative Factors under an Inductive Bias of Confounder

arXiv: https://arxiv.org/abs/2310.17325 · HTML: https://mlsanity.com/paper/2310.17325/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 27 references that are in the map.
- First posted on arXiv 26 Oct 2023.

## Builds on (most cited first, 16 of 16)

- [Deep Learning Face Attributes in the Wild](https://mlsanity.com/paper/1411.7766/) (2014)
- [Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations](https://mlsanity.com/paper/1811.12359/) (2018)
- [InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets](https://mlsanity.com/paper/1606.03657/) (2016)
- [Disentangling by Factorising](https://mlsanity.com/paper/1802.05983/) (2018)
- [Weakly-Supervised Disentanglement Without Compromises](https://mlsanity.com/paper/2002.02886/) (2020)
- [Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoders](https://mlsanity.com/paper/1611.02648/) (2016)
- [Weakly supervised causal representation learning](https://mlsanity.com/paper/2203.16437/) (2022)
- [Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustness](https://mlsanity.com/paper/1811.00007/) (2018)
- [Desiderata for Representation Learning: A Causal Perspective](https://mlsanity.com/paper/2109.03795/) (2021)
- [Weakly Supervised Representation Learning with Sparse Perturbations](https://mlsanity.com/paper/2206.01101/) (2022)
- [Equitability, mutual information, and the maximal information coefficient](https://mlsanity.com/paper/1301.7745/) (2013)
- [Weakly Supervised Disentanglement by Pairwise Similarities](https://mlsanity.com/paper/1906.01044/) (2019)
- [Understanding the Generalization Benefit of Model Invariance from a Data Perspective](https://mlsanity.com/paper/2111.05529/) (2021)
- [On Causally Disentangled Representations](https://mlsanity.com/paper/2112.05746/) (2021)
- [CAT: Causal Audio Transformer for Audio Classification](https://mlsanity.com/paper/2303.07626/) (2023)
- [PatchMix Augmentation to Identify Causal Features in Few-shot Learning](https://mlsanity.com/paper/2211.16019/) (2022)

## Built on by (most cited first, 7 of 7)

- [Mementos: A Comprehensive Benchmark for Multimodal Large Language Model Reasoning over Image Sequences](https://mlsanity.com/paper/2401.10529/) (2024)
- [Explore Spurious Correlations at the Concept Level in Language Models for Text Classification](https://mlsanity.com/paper/2311.08648/) (2023)
- [Multi-Stage Balanced Distillation: Addressing Long-Tail Challenges in Sequence-Level Knowledge Distillation](https://mlsanity.com/paper/2406.13114/) (2024)
- [Grouped Discrete Representation for Object-Centric Learning](https://mlsanity.com/paper/2411.02299/) (2024)
- [CSRec: Rethinking Sequential Recommendation from A Causal Perspective](https://mlsanity.com/paper/2409.05872/) (2024)
- [Weakly Augmented Variational Autoencoder in Time Series Anomaly Detection](https://mlsanity.com/paper/2401.03341/) (2024)
- [FlexCausal: Flexible Causal Disentanglement via Structural Flow Priors and Manifold-Aware Interventions](https://mlsanity.com/paper/2601.21567/) (2026)

## Abstract (arXiv)

Representation learning assumes that real-world data is generated by a few semantically meaningful generative factors (i.e., sources of variation) and aims to discover them in the latent space. These factors are expected to be causally disentangled, meaning that distinct factors are encoded into separate latent variables, and changes in one factor will not affect the values of the others. Compared to statistical independence, causal disentanglement allows more controllable data generation, improved robustness, and better generalization. However, most existing work assumes unconfoundedness in the discovery process, that there are no common causes to the generative factors and thus obtain only statistical independence. In this paper, we recognize the importance of modeling confounders in discovering causal generative factors. Unfortunately, such factors are not identifiable without proper inductive bias. We fill the gap by introducing a framework entitled Confounded-Disentanglement (C-Disentanglement), the first framework that explicitly introduces the inductive bias of confounder via labels from domain expertise. In addition, we accordingly propose an approach to sufficiently identify the causally disentangled factors under any inductive bias of the confounder. We conduct extensive experiments on both synthetic and real-world datasets. Our method demonstrates competitive results compared to various SOTA baselines in obtaining causally disentangled features and downstream tasks under domain shifts.