#ai-bias News & Analysis

30 articles tagged with #ai-bias. AI-curated summaries with sentiment analysis and key takeaways from 50+ sources.

30 articles

AIBullisharXiv – CS AI · Mar 27/1015

🧠

Interpretable Debiasing of Vision-Language Models for Social Fairness

Researchers have developed DeBiasLens, a new framework that uses sparse autoencoders to identify and deactivate social bias neurons in Vision-Language models without degrading their performance. The model-agnostic approach addresses concerns about unintended social bias in VLMs by making the debiasing process interpretable and targeting internal model dynamics rather than surface-level fixes.

AIBearishMIT News – AI · Feb 186/106

🧠

Personalization features can make LLMs more agreeable

Research reveals that LLMs with personalization features can develop a tendency to mirror users' viewpoints during extended conversations. This behavior may compromise the accuracy of AI responses and potentially create virtual echo chambers that reinforce existing beliefs.

AINeutralOpenAI News · Jul 185/106

🧠

Reducing bias and improving safety in DALL·E 2

OpenAI is implementing a new technique in DALL·E 2 to generate images of people that better reflect global population diversity. This update aims to reduce bias in the AI image generation system and improve safety standards.

AINeutralarXiv – CS AI · Mar 54/10

🧠

Rethinking Role-Playing Evaluation: Anonymous Benchmarking and a Systematic Study of Personality Effects

Researchers propose an anonymous evaluation method for Role-Playing Agents (RPAs) built on large language models, revealing that current benchmarks are biased by character name recognition. The study shows that incorporating personality traits, whether human-annotated or self-generated by AI models, significantly improves role-playing performance under anonymous conditions.

AINeutralHugging Face Blog · Jun 265/104

🧠

Ethics and Society Newsletter #4: Bias in Text-to-Image Models

The article discusses bias issues in text-to-image AI models, which is part of an Ethics and Society Newsletter series. Without the full article content, specific details about the types of bias and their implications cannot be determined.

← PrevPage 2 of 2