Bias Amplification

Bias amplification, the phenomenon where machine learning models exacerbate existing biases in training data, is a critical area of research aiming to understand and mitigate unfair outcomes in various applications. Current investigations focus on identifying bias amplification across diverse model architectures, including graph neural networks, diffusion models, and recommendation systems, and exploring mitigation strategies such as re-weighting samples, balanced message passing, and fairness-aware sampling. Understanding and addressing bias amplification is crucial for ensuring fairness and equity in AI systems, impacting fields ranging from loan applications and hiring processes to image generation and social media algorithms. The ultimate goal is to develop models that are both accurate and unbiased.

Papers

October 19, 2024

Bias Amplification: Language Models as Increasingly Biased Media
Ze Wang, Zekun Wu, Jeremy Zhang, Navya Jain, Xin Guan, Adriano Koshiyama
Language Model Synthetic Data Absolute Stance Bias Medium Bias Bias Amplification

October 7, 2024

An Effective Theory of Bias Amplification
Arjun Subramonian, Samuel J. Bell, Levent Sagun, Elvis Dohmatob
Machine Learning Model Absolute Stance Bias Ridge Regression Bias Amplification Effective Theory

June 6, 2024

Mitigating Bias in Dataset Distillation
Justin Cui, Ruochen Wang, Yuanhao Xiong, Cho-Jui Hsieh
Synthetic Data Dataset Condensation Bias Amplification Background Bias

May 24, 2024

Mind the Gap: A Causal Perspective on Bias Amplification in Prediction & Decision-Making
Drago Plecko, Elias Bareinboim
Human Prediction Decision Making Fair Machine Learning Fairness Criterion Bias Amplification Causal Concept

April 17, 2024

Clipped SGD Algorithms for Privacy Preserving Performative Prediction: Bias Amplification and Remedies
Qiang Li, Michal Yemini, Hoi-To Wai
Optimization Purpose Stochastic Gradient Descent Performative Prediction Clipped Stochastic Gradient Descent Bias Amplification Fast AltGDA Type Algorithm Stable Optimization

January 6, 2024

Fair Sampling in Diffusion Models through Switching Mechanism
Yujin Choi, Jinseong Park, Hoki Kim, Jaewook Lee, Saerom Park
Diffusion Model Functional Mechanism Bias Amplification Statistical Fairness FAIR Data

December 8, 2023

Understanding Community Bias Amplification in Graph Representation Learning
Shengzhong Zhang, Wenjie Yang, Yimin Zhang, Hongwei Zhang, Divin Yan, Zengfeng Huang
Contrastive Learning Graph Representation Learning Graph Contrastive Learning Graph Coarsening Bias Amplification

November 10, 2023

DPR: An Algorithm Mitigate Bias Accumulation in Recommendation feedback loops
Hangtong Xu, Yuanbo Xu, Yongjian Yang, Fuzhen Zhuang, Hui Xiong
Recommendation System Recommendation Model Negative Sampling Recommendation Quality Bias Amplification Bias Mitigation Algorithm

October 26, 2023

Evaluating Bias and Fairness in Gender-Neutral Pretrained Vision-and-Language Models
Laura Cabello, Emanuele Bugliarello, Stephanie Brandl, Desmond Elliott
Procedural Fairness Gender Bias Pre Trained Vision Language Model Bias Amplification Evaluating Bias

August 1, 2023

The Bias Amplification Paradox in Text-to-Image Generation
Preethi Seshadri, Sameer Singh, Yanai Elazar
Text to Image Generation Gender Bias Bias Measurement Bias Amplification

June 7, 2023

BeMap: Balanced Message Passing for Fair Graph Neural Network
Xiao Lin, Jian Kang, Weilin Cong, Hanghang Tong
Graph Neural Network Procedural Fairness Absolute Stance Bias Node Classification Message Passing Bias Amplification Balanced Sampling

January 18, 2023

A Comparative Analysis of Bias Amplification in Graph Neural Network Approaches for Recommender Systems
Nikzad Chizari, Niloufar Shoeibi, María N. Moreno-García
Graph Neural Network Recommender System Comparative Study Matrix Factorization GNN Based Bias Amplification

October 21, 2022

Men Also Do Laundry: Multi-Attribute Bias Amplification
Dora Zhao, Jerone T. A. Andrews, Alice Xiang
Gender Bias Bias Mitigation Bias Amplification

August 15, 2022

Bias amplification in experimental social networks is reduced by resampling
Mathew D. Hardy, Bill D. Thompson, P. M. Krafft, Thomas L. Griffiths
Social Network Biased Decision Social Influence Bias Amplification

May 29, 2022

Revisiting the Importance of Amplifying Bias for Debiasing
Jungsoo Lee, Jeonghoon Park, Daeyoung Kim, Juyoung Lee, Edward Choi, Jaegul Choo
Importance Aware Self Debiasing Debiasing Method Dataset Bias Biased Datasets Bias Amplification Bias Aligned Sample

March 29, 2022

Quantifying Societal Bias Amplification in Image Captioning
Yusuke Hirota, Yuta Nakashima, Noa Garcia
Image Captioning Bias Mitigation Bias Amplification

January 27, 2022

A Systematic Study of Bias Amplification
Melissa Hall, Laurens van der Maaten, Laura Gustafson, Maxwell Jones, Aaron Adcock
Machine Learning Model Systematic Study Bias Amplification Strong Model