Latest Articles for Bias

Introducing autophagy penalized likelihood estimation to address bias in AI models.

EquiPrompt aims to reduce biases in AI-generated images using innovative methods.

Evaluating risks of biased outcomes in robots using language models.

Study reveals gender bias in chatGPT's career suggestions for children.

Addressing bias in source localization to enhance accuracy across various fields.

Examining the impact of dialects on language bias detection accuracy.

Study reveals biases in AI hiring recommendations based on candidate names.

Explore the principles and applications of stochastic approximation in uncertain environments.

A study highlights discrepancies in AI access among ethnic groups.

Examining biases in language models used for mental health analysis and solutions.

Examining how cultural bias affects AI image understanding.

Examining the roots and implications of bias in language technology.

Study analyzes generalization and performance of random feature ridge regression using eigenvalues.

Examining the issues and potential improvements in academic peer review.

A study on using LLMs to judge other LLMs and its implications.

This study analyzes the roles of sources in Twitter's Community Notes.

Exploring techniques for reducing bias in advanced language models.

A new tool to evaluate biases in large vision-language models.

New approaches enhance PCR accuracy in DNA analysis.

New method addresses regional differences in gender bias evaluation.

A novel framework enhances graph matching without labeled data.

This article reviews tokenization issues and proposes solutions for bias reduction.

Examining the role of explainable AI in addressing bias.

This article examines how bias develops during the training of machine learning models.

FineWeb offers 15 trillion tokens to improve language model training.

Study shows LLMs favor native speakers, raising equality concerns.

This study investigates biases in Bangla language models and their societal effects.

A new method enhances power spectrum estimation from nonuniformly sampled data.

Researchers use propositional probes to enhance the reliability of language models.

Examining the need for fairness in AI and its impact on society.

This workshop analyzes gender stereotypes in AI through the lens of societal biases.

A study on bias in synthetic image generation models and their societal impacts.

PopAlign aims to reduce bias in text-to-image generation while maintaining quality.

Exploring the challenges and solutions for bias in machine learning.

A novel method enhances the accuracy of treatment effect estimates over time.

Examining the relationship between privacy techniques and biases in language models.

Tools like OxonFair help ensure fairness in AI decision-making.

Names from different countries impact how classifiers interpret social media content.

A framework to reduce bias in AI language models while maintaining accuracy.

Exploring how synthetic data shapes machine learning models and their behavior.