Bias Discovery

Bias discovery in machine learning focuses on identifying and understanding how models unfairly discriminate against certain groups, hindering their performance and potentially causing harm. Current research emphasizes detecting biases in various model architectures, including large vision-language models and those used in mental health applications, employing techniques like counterfactual analysis, latent space exploration, and keyword-based explanations to uncover spurious correlations within data. This work is crucial for improving model fairness, robustness, and trustworthiness across diverse applications, ultimately leading to more equitable and reliable AI systems.

Papers

July 2, 2024

ViG-Bias: Visually Grounded Bias Discovery and Mitigation
Badr-Eddine Marani, Mohamed Hanini, Nihitha Malayarukil, Stergios Christodoulidis, Maria Vakalopoulou, Enzo Ferrante
Faulty Negative Mitigation Visual Explanation Inherent Bias Modal Embeddings Bias Discovery

June 5, 2024

Discovering Bias in Latent Space: An Unsupervised Debiasing Approach
Dyah Adila, Shuai Zhang, Boran Han, Yuyang Wang
Latent Space Supervised Baseline Instruction Tuned Model Debiasing Framework Correctness Check Bias Discovery

May 30, 2024

Uncovering Bias in Large Vision-Language Models at Scale with Counterfactuals
Phillip Howard, Kathleen C. Fraser, Anahita Bhiwandiwalla, Svetlana Kiritchenko
Large Vision Language Model Visual Analogue Scale Social Bias High Quality Counterfactuals Visual Input Bias Discovery

March 29, 2024

Uncovering Bias in Large Vision-Language Models with Counterfactuals
Phillip Howard, Anahita Bhiwandiwalla, Kathleen C. Fraser, Svetlana Kiritchenko
Large Vision Language Model Social Bias High Quality Counterfactuals Visual Input Multimodal Dialogue Bias Discovery

January 24, 2024

Common-Sense Bias Discovery and Mitigation for Classification Tasks
Miao Zhang, Zee fryer, Ben Colman, Ali Shahriyari, Gaurav Bharaj
Classification Task Model Bias Faulty Negative Mitigation Feature Correlation Benchmark Image Bias Discovery Downstream Bias

September 29, 2023

FACTS: First Amplify Correlations and Then Slice to Discover Bias
Sriram Yenamandra, Pratik Ramesh, Viraj Prabhu, Judy Hoffman
Model Slicing Slice by Slice Bias Related Issue Space Bias Bias Discovery Correlation Bias

August 18, 2023

Data augmentation and explainability for bias discovery and mitigation in deep learning
Agnieszka Mikołajczyk-Bareła
Deep Learning Data Augmentation High Explainability Faulty Negative Mitigation Bias Detection Texture Bias Bias Discovery

May 12, 2023

Surfacing Biases in Large Language Models using Contrastive Input Decoding
Gal Yona, Or Honovich, Itay Laish, Roee Aharoni
Large Language Model Faithful Generation Contextual Biasing Contrastive Decoding Bias Discovery Open Ended Natural Language Generation

March 27, 2023

Uncovering Bias in Personal Informatics
Sofia Yfantidou, Pavlos Sermpezis, Athena Vakali, Ricardo Baeza-Yates
Absolute Stance Bias Wearable Device Intersectional Bias Bias Discovery Data Bias

February 22, 2023

Uncovering Bias in Face Generation Models
Cristian Muñoz, Sara Zannone, Umar Mohammed, Adriano Koshiyama
GAN Model Bias Mitigation Group Fairness Face Generation Feature Perturbation Bias Discovery

January 26, 2023

Discovering and Mitigating Visual Biases through Keyword Explanation
Younghyun Kim, Sangwoo Mo, Minkyu Kim, Kyungmin Lee, Jaeho Lee, Jinwoo Shin
Scientific Discovery Contextual Biasing Keyword Enrollment Bias Discovery Visual Bias

May 24, 2022

Bias Discovery in Machine Learning Models for Mental Health
Pablo Mosteiro, Jesse Kuiper, Judith Masthoff, Floortje Scheepers, Marco Spruit
Machine Learning Model Absolute Stance Bias Mental Health Bias Mitigation Fairness Metric AI Fairness Bias Discovery