# Information-Theoretic Testing and Debugging of Fairness Defects in Deep Neural Networks

arXiv: https://arxiv.org/abs/2304.04199 · HTML: https://mlsanity.com/paper/2304.04199/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 13 of its 58 references that are in the map.
- First posted on arXiv 9 Apr 2023.

## Builds on (most cited first, 13 of 13)

- [Equality of Opportunity in Supervised Learning](https://mlsanity.com/paper/1610.02413/) (2016)
- [Mitigating Unwanted Biases with Adversarial Learning](https://mlsanity.com/paper/1801.07593/) (2018)
- [A Reductions Approach to Fair Classification](https://mlsanity.com/paper/1803.02453/) (2018)
- [Fairness Testing: Testing Software for Discrimination](https://mlsanity.com/paper/1709.03221/) (2017)
- [Bias in Machine Learning Software: Why? How? What to do?](https://mlsanity.com/paper/2105.12195/) (2021)
- [Automated Directed Fairness Testing](https://mlsanity.com/paper/1807.00468/) (2018)
- [NeuronFair: Interpretable White-Box Fairness Testing through Biased Neuron Identification](https://mlsanity.com/paper/2112.13214/) (2021)
- [Finding Faster Configurations using FLASH](https://mlsanity.com/paper/1801.02175/) (2018)
- [Fairness-aware Configuration of Machine Learning Libraries](https://mlsanity.com/paper/2202.06196/) (2022)
- [Automated Test Generation to Detect Individual Discrimination in AI Models](https://mlsanity.com/paper/1809.03260/) (2018)
- [Algorithmic Fairness and Vertical Equity: Income Fairness with IRS Tax Audit Models](https://mlsanity.com/paper/2206.09875/) (2022)
- [Metamorphic Testing and Debugging of Tax Preparation Software](https://mlsanity.com/paper/2205.04998/) (2022)
- [QFuzz: Quantitative Fuzzing for Side Channels](https://mlsanity.com/paper/2106.03346/) (2021)

## Built on by (most cited first, 10 of 10)

- [Beyond Deepfake Images: Detecting AI-Generated Videos](https://mlsanity.com/paper/2404.15955/) (2024)
- [Causality-Aided Trade-off Analysis for Machine Learning Fairness](https://mlsanity.com/paper/2305.13057/) (2023)
- [FairLay-ML: Intuitive Debugging of Fairness in Data-Driven Social-Critical Software](https://mlsanity.com/paper/2407.01423/) (2024)
- [FairQuant: Certifying and Quantifying Fairness of Deep Neural Networks](https://mlsanity.com/paper/2409.03220/) (2024)
- [A Novel Approach to Chest X-ray Lung Segmentation Using U-net and Modified Convolutional Block Attention Module](https://mlsanity.com/paper/2404.14322/) (2024)
- [FairSense: Long-Term Fairness Analysis of ML-Enabled Systems](https://mlsanity.com/paper/2501.01665/) (2025)
- [Hierarchical SegNet with Channel and Context Attention for Accurate Lung Segmentation in Chest X-ray Images](https://mlsanity.com/paper/2405.12318/) (2024)
- [Unbiasing on the Fly: Explanation-Guided Human Oversight of Machine Learning System Decisions](https://mlsanity.com/paper/2406.17906/) (2024)
- [Uncovering Discrimination Clusters: Quantifying and Explaining Systematic Fairness Violations](https://mlsanity.com/paper/2512.23769/) (2025)
- [Analyzing Fairness of Neural Network Prediction via Counterfactual Dataset Generation](https://mlsanity.com/paper/2602.10457/) (2026)

## Abstract (arXiv)

The deep feedforward neural networks (DNNs) are increasingly deployed in socioeconomic critical decision support software systems. DNNs are exceptionally good at finding minimal, sufficient statistical patterns within their training data. Consequently, DNNs may learn to encode decisions -- amplifying existing biases or introducing new ones -- that may disadvantage protected individuals/groups and may stand to violate legal protections. While the existing search based software testing approaches have been effective in discovering fairness defects, they do not supplement these defects with debugging aids -- such as severity and causal explanations -- crucial to help developers triage and decide on the next course of action. Can we measure the severity of fairness defects in DNNs? Are these defects symptomatic of improper training or they merely reflect biases present in the training data? To answer such questions, we present DICE: an information-theoretic testing and debugging framework to discover and localize fairness defects in DNNs. The key goal of DICE is to assist software developers in triaging fairness defects by ordering them by their severity. Towards this goal, we quantify fairness in terms of protected information (in bits) used in decision making. A quantitative view of fairness defects not only helps in ordering these defects, our empirical evaluation shows that it improves the search efficiency due to resulting smoothness of the search space. Guided by the quantitative fairness, we present a causal debugging framework to localize inadequately trained layers and neurons responsible for fairness defects. Our experiments over ten DNNs, developed for socially critical tasks, show that DICE efficiently characterizes the amounts of discrimination, effectively generates discriminatory instances, and localizes layers/neurons with significant biases.