Fairness Certification

Fairness certification aims to provide verifiable guarantees that machine learning models do not discriminate against protected groups, addressing growing concerns about algorithmic bias in high-stakes applications. Current research focuses on developing methods to certify fairness for various model architectures, including neural networks and k-nearest neighbors, often employing techniques like zero-knowledge proofs, formal verification (e.g., SMT-based approaches), and statistical bounds on unfairness. This field is crucial for building trust in AI systems and ensuring equitable outcomes across diverse populations, impacting both the development of robust fairness metrics and the deployment of responsible AI in society.

Papers

June 7, 2024

CTBENCH: A Library and Benchmark for Certified Training
Yuhao Mao, Stefan Balauca, Martin Vechev
Training Data New Benchmark Hyperparameter Tuning Benchmark Study Easy to Use Library Robust Neural Network Fairness Certification

February 19, 2024

FairProof : Confidential and Certifiable Fairness for Neural Networks
Chhavi Yadav, Amrita Roy Chowdhury, Dan Boneh, Kamalika Chaudhuri
Neural Network Procedural Fairness Zero Knowledge Fairness Property Fairness Certification

January 2, 2024

Fairness Certification for Natural Language Processing and Large Language Models
Vincent Freiberger, Erik Buchmann
Natural Language Processing Algorithmic Fairness Fairness Certification

July 17, 2023

Certifying the Fairness of KNN in the Presence of Dataset Bias
Yannan Li, Jingbo Wang, Chao Wang
Procedural Fairness Speech Presence Dataset Bias K Nearest Neighbor Fairness Certification

December 8, 2022

Fairify: Fairness Verification of Neural Networks
Sumon Biswas, Hridesh Rajan
Neural Network Fairness Property Fairness Certification Fairness Verification

December 5, 2022

Certifying Fairness of Probabilistic Circuits
Nikil Roashan Selvam, Guy Van den Broeck, YooJung Choi
Probabilistic Model Algorithmic Fairness Probabilistic Circuit Fairness Certification Non Discrimination

October 13, 2022

FARE: Provably Fair Representation Learning with Practical Certificates
Nikola Jovanović, Mislav Balunović, Dimitar I. Dimitrov, Martin Vechev
Jina Embeddings Fairness Accuracy FAir Classifier Fair Representation Learning Fairness Certification Sample Wise Certificate

July 20, 2022

Measuring and signing fairness as performance under multiple stakeholder distributions
David Lopez-Paz, Diane Bouchacourt, Levent Sagun, Nicolas Usunier
Procedural Fairness System Performance Fairness Metric Biased Prediction Multi Stakeholder Fairness Property Fairness Certification

May 31, 2022

Certifying Some Distributional Fairness with Subpopulation Decomposition
Mintong Kang, Linyi Li, Maurice Weber, Yang Liu, Ce Zhang, Bo Li
Machine Learning Model Distributional Robustness Fairness Certification Semantically Meaningful Subpopulation Property Professional Certification

January 10, 2022

Fairness Score and Process Standardization: Framework for Fairness Certification in Artificial Intelligence Systems
Avinash Agarwal, Harsh Agarwal, Nihaarika Agarwal
New Framework Artificial Intelligence System Data Standardization Fairness Performance Fairness Certification

November 26, 2021

Latent Space Smoothing for Individually Fair Representations
Momchil Peychev, Anian Ruoss, Mislav Balunović, Maximilian Baader, Martin Vechev
Latent Space Generative Modeling Fair Representation Latent Space Optimization Fair Representation Learning Fairness Certification