# Null-sampling for Interpretable and Fair Representations

arXiv: https://arxiv.org/abs/2008.05248 · HTML: https://mlsanity.com/paper/2008.05248/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Aug 2020.

## Built on by (most cited first, 16 of 16)

- [Fair Normalizing Flows](https://mlsanity.com/paper/2106.05937/) (2021)
- [Fair Representation Learning through Implicit Path Alignment](https://mlsanity.com/paper/2205.13316/) (2022)
- [Fairness and Bias in Multimodal AI: A Survey](https://mlsanity.com/paper/2406.19097/) (2024)
- [Latent Space Smoothing for Individually Fair Representations](https://mlsanity.com/paper/2111.13650/) (2021)
- [Fair Representation: Guaranteeing Approximate Multiple Group Fairness for Unknown Tasks](https://mlsanity.com/paper/2109.00545/) (2021)
- [A Critical Survey on Fairness Benefits of Explainable AI](https://mlsanity.com/paper/2310.13007/) (2023)
- [Certifying Some Distributional Fairness with Subpopulation Decomposition](https://mlsanity.com/paper/2205.15494/) (2022)
- [Towards Group Robustness in the presence of Partial Group Labels](https://mlsanity.com/paper/2201.03668/) (2022)
- [A Large-Scale Empirical Study on Improving the Fairness of Image Classification Models](https://mlsanity.com/paper/2401.03695/) (2024)
- [Personalizing Pre-trained Models](https://mlsanity.com/paper/2106.01499/) (2021)
- [RealPatch: A Statistical Matching Framework for Model Patching with Real Samples](https://mlsanity.com/paper/2208.02192/) (2022)
- [Okapi: Generalising Better by Making Statistical Matches Match](https://mlsanity.com/paper/2211.05236/) (2022)
- [A Survey on Preserving Fairness Guarantees in Changing Environments](https://mlsanity.com/paper/2211.07530/) (2022)
- [Addressing Missing Sources with Adversarial Support-Matching](https://mlsanity.com/paper/2203.13154/) (2022)
- [Constructing Fair Latent Space for Intersection of Fairness and Explainability](https://mlsanity.com/paper/2412.17523/) (2024)
- [Fairness meets Cross-Domain Learning: a new perspective on Models and Metrics](https://mlsanity.com/paper/2303.14411/) (2023)

## Abstract (arXiv)

We propose to learn invariant representations, in the data domain, to achieve interpretability in algorithmic fairness. Invariance implies a selectivity for high level, relevant correlations w.r.t. class label annotations, and a robustness to irrelevant correlations with protected characteristics such as race or gender. We introduce a non-trivial setup in which the training set exhibits a strong bias such that class label annotations are irrelevant and spurious correlations cannot be distinguished. To address this problem, we introduce an adversarially trained model with a null-sampling procedure to produce invariant representations in the data domain. To enable disentanglement, a partially-labelled representative set is used. By placing the representations into the data domain, the changes made by the model are easily examinable by human auditors. We show the effectiveness of our method on both image and tabular datasets: Coloured MNIST, the CelebA and the Adult dataset.