Inherent Bias

Inherent bias in artificial intelligence systems, stemming from biased training data and algorithmic design, is a significant concern, with research focusing on its detection and mitigation. Current efforts involve developing novel methods for data partitioning to minimize accidental bias in controlled trials, employing visual explanations to improve bias discovery in machine learning models, and using techniques like counterfactual debating and prompt engineering to control bias in large language models and generative models. Addressing inherent bias is crucial for ensuring fairness, accuracy, and trustworthiness in AI applications across various domains, impacting both the development of responsible AI and the ethical implications of its deployment.

Papers

October 13, 2023

Towards Informative Few-Shot Prompt with Maximum Information Gain for In-Context Learning
Hongfu Liu, Ye Wang
Large Language Model Context Learning Style PROMPT Information Gain Inherent Bias Learning Paradigm

September 25, 2023

Noise-in, Bias-out: Balanced and Real-time MoCap Solving
Georgios Albanis, Nikolaos Zioulis, Spyridon Thermos, Anargyros Chatzitofis, Kostas Kolomvatsos
Inherent Bias Noise Injection Balanced Approach Marker Less Mocap Optical Motion Capture Marker Selection

September 7, 2023

OpinionGPT: Modelling Explicit Biases in Instruction-Tuned LLMs
Patrick Haller, Ansar Aynetdinov, Alan Akbik
Instruction Tuning Human Opinion Instruction Tuned Large Language Model Inherent Bias Potential Bias

August 21, 2023

Systematic Offensive Stereotyping (SOS) Bias in Language Models
Fatma Elsafoury
Language Model Absolute Stance Bias Hate Speech Detection Debiasing Method Online Hate Inherent Bias

July 24, 2023

Interpretable Stereotype Identification through Reasoning
Jacob-Junqi Tian, Omkar Dige, David Emerson, Faiza Khan Khattak
Language Model Zero Shot Complex Reasoning Stereotype Content Inherent Bias Stereotype Detection

May 3, 2023

Fairness in AI Systems: Mitigating gender bias from language-vision models
Lavisha Aggarwal, Shruti Bhargava
Procedural Fairness Gender Bias AI System Inherent Bias Biased Data Language Vision Model

April 13, 2023

Evaluation of Social Biases in Recent Large Pre-Trained Models
Swapnil Sharma, Nikita Anand, Kranthi Kiran G. V., Alind Jain
Global Evaluation Social Bias Large Pre Trained Model Large Pre Trained Language Model Inherent Bias Bias Reduction

February 22, 2023

Delving into Identify-Emphasize Paradigm for Combating Unknown Bias
Bowen Zhao, Chen Chen, Qian-Wei Wang, Anfeng He, Shu-Tao Xia
Task Specific Model Dataset Bias Inherent Bias Bias Score Bias Aligned Sample

December 13, 2022

Fair Infinitesimal Jackknife: Mitigating the Influence of Biased Training Data Points Without Refitting
Prasanna Sattigeri, Soumya Ghosh, Inkit Padhi, Pierre Dognin, Kush R. Varshney
Machine Learning Model Procedural Fairness External Influence Inherent Bias Biased Training Balancing Fairness Pre Trained Classifier

August 29, 2022

Towards Explaining Demographic Bias through the Eyes of Face Recognition Models
Biying Fu, Naser Damer
Face Recognition Model Inherent Bias Biased Behavior Character Eye Demographic Bias Human Preference Judgment

June 29, 2022

Identifying and Combating Bias in Segmentation Networks by leveraging multiple resolutions
Leonie Henschel, David Kügler, Derek S Andrews, Christine W Nordahl, Martin Reuter
Absolute Stance Bias Segmentation Network Multi Resolution Low Resolution Inherent Bias Image Resolution Training Resolution

June 20, 2022

Fewer Errors, but More Stereotypes? The Effect of Model Size on Gender Bias
Yarden Tal, Inbal Magar, Roy Schwartz
Language Model Mixed Effect Gender Bias Error Detection Gender Information Model Size Inherent Bias

June 9, 2022

Unlearning Protected User Attributes in Recommendations with Adversarial Training
Christian Ganhör, David Penz, Navid Rekabsaz, Oleg Lesota, Markus Schedl
Adversarial Training Bias Mitigation Movie Recommendation Recommendation Performance Implicit Knowledge Inherent Bias Encode Bias

June 8, 2022

Improving Evaluation of Debiasing in Image Classification
Jungsoo Lee, Juyoung Lee, Sanghun Jung, Jaegul Choo
Image Classification Self Debiasing Dataset Bias Inherent Bias Bias Attribute Bias Aligned Sample

May 26, 2022

DeepTechnome: Mitigating Unknown Bias in Deep Learning Based Assessment of CT Images
Simon Langer, Oliver Taubmann, Felix Denzinger, Andreas Maier, Alexander Mühlberg
Bias Mitigation CT Image Technology Information Data Inherent Bias Adversarial Testing Biased Feature Based Assessment

May 10, 2022

Social Inclusion in Curated Contexts: Insights from Museum Practices
Han-Yin Huang, Cynthia C. S. Liem
Artificial Intelligence Context Information DCU Insight AQ Inherent Bias Diverse Population Inclusive Approach

May 9, 2022

Counterfactually Augmented Data and Unintended Bias: The Case of Sexism and Hate Speech Detection
Indira Sen, Mattia Samory, Claudia Wagner, Isabelle Augenstein
Case Relevance Hate Speech Detection Model Bias Model Robustness Augmented Data Inherent Bias Domain Generalizability

January 27, 2022

FairEGM: Fair Link Prediction and Recommendation via Emulated Graph Modification
Sean Current, Yuntian He, Saket Gurukar, Srinivasan Parthasarathy
Graph Neural Network Link Prediction Effective Recommendation Inherent Bias Recommendation Fairness Graph Injection Attack Fairness Optimization

January 17, 2022

Unintended Bias in Language Model-driven Conversational Recommendation
Tianshu Shen, Jiaru Li, Mohamed Reda Bouadjenek, Zheda Mai, Scott Sanner
Language Model Inherent Bias Recommendation Policy Conversational Recommendation System

November 25, 2021

Combating Unknown Bias with Effective Bias-Conflicting Scoring and Gradient Alignment
Bowen Zhao, Chen Chen, Qian-Wei Wang, Anfeng He, Shu-Tao Xia
Task Specific Model Dataset Bias Inherent Bias Gradient Alignment Bias Score Bias Aligned Sample