Bias Detection

Bias detection in artificial intelligence focuses on identifying and mitigating unfairness stemming from biases present in training data and model architectures. Current research emphasizes developing robust and explainable methods for detecting biases across diverse applications, including language models, image generation, and medical decision-making, often leveraging transformer-based models and techniques like anomaly detection or contrastive learning. This work is crucial for ensuring fairness and ethical considerations in AI systems, impacting both the development of more equitable algorithms and the responsible deployment of AI in sensitive domains like healthcare and legal systems.

Papers

May 22, 2023

This Prompt is Measuring <MASK>: Evaluating Bias Evaluation in Language Models
Seraphina Goldfarb-Tarrant, Eddie Ungless, Esma Balkir, Su Lin Blodgett
Language Model NLP Field Style PROMPT Social Bias Bias Evaluation Mask Frozen DETR Bias Detection Publication Bias

February 14, 2023

BiasTestGPT: Using ChatGPT for Social Bias Testing of Language Models
Rafal Kocielnik, Shrimai Prabhumoye, Vivian Zhang, Roy Jiang, R. Michael Alvarez, Anima Anandkumar
Language Model ChatGPT Generated Conversation Pretrained Language Model Social Bias Bias Detection Sentence Generation Bias Testing

December 20, 2022

Trustworthy Social Bias Measurement
Rishi Bommasani, Percy Liang
Social Bias Bias Detection Prior Model Bias Measurement

November 16, 2022

Auditing Algorithmic Fairness in Machine Learning for Health with Severity-Based LOGAN
Anaelia Ovalle, Sunipa Dev, Jieyu Zhao, Majid Sarrafzadeh, Kai-Wei Chang
Machine Learning Bias Detection Severity Classification Machine Learning Fairness Fairness Audit

October 13, 2022

SODAPOP: Open-Ended Discovery of Social Biases in Social Commonsense Reasoning Models
Haozhe An, Zongxia Li, Jieyu Zhao, Rachel Rudinger
Scientific Discovery Commonsense Knowledge Social Bias Bias Detection Social Commonsense

October 11, 2022

Race Bias Analysis of Bona Fide Errors in face anti-spoofing
Latifah Abduh, Ioannis Ivrissimtzis
Absolute Stance Bias Human Face Anti Spoofing Human Answer Mistake Racial Bias Bias Detection Potential Bias

June 17, 2022

BITS Pilani at HinglishEval: Quality Evaluation for Code-Mixed Hinglish Text Using Transformers
Shaz Furniturewala, Vijay Kumari, Amulya Ratna Dash, Hriday Kedia, Yashvardhan Sharma
Transformer Megatron Decepticons Quality Assessment Multilingual BERT Code Mixed Text Data Bias Detection

April 28, 2022

Learning to Split for Automatic Bias Detection
Yujia Bao, Regina Barzilay
LeArning Abstract Biased Datasets Split and Fit Human Bias Bias Detection Debiasing Framework Optimal Split

March 24, 2022

Intrinsic Bias Identification on Medical Image Datasets
Shijie Zhang, Lanjun Wang, Lian Ding, An-an Liu, Senhua Zhu, Dandan Tu
Medical Image Datasets Bias Metric Bias Detection Intrinsic Bias Bias Attribute

January 25, 2022

Modeling Multi-level Context for Informational Bias Detection by Contrastive Learning and Sentential Graph Network
Shijia Guo, Kenny Q. Zhu
Contrastive Learning Sentence Level Biased Behavior Bias Detection Sentence Entity Graph

December 15, 2021

Few-shot Instruction Prompts for Pretrained Language Models to Detect Social Biases
Shrimai Prabhumoye, Rafal Kocielnik, Mohammad Shoeybi, Anima Anandkumar, Bryan Catanzaro
Pre Trained Language Model Pretrained Language Model Social Bias Shot Prompting Bias Detection Fine Grained Bias

December 14, 2021

Towards A Reliable Ground-Truth For Biased Language Detection
Timo Spinde, David Krieger, Manuel Plank, Bela Gipp
Ground Truth Biased Behavior Language Bias Bias Detection Crowdsourced Label Crowdsourcing Model