Find a story
Search Spins
Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.
5 results for “fairness”
The Fairness Collapse Phenomenon: Bias Amplification in Language Models Trained on Synthetic Data
Researchers identify a new phenomenon—'fairness collapse'—where language models trained recursively on synthetic data amplify social biases faster than they degrade in standard performance metrics, posing a stealth risk to AI equity.
Aug 6, 2026
Warren demands investigation into Trump meme coin
U.S. Senators Elizabeth Warren and Richard Blumenthal formally requested the SEC investigate a cryptocurrency token associated with Donald Trump, citing concerns it may constitute securities fraud or a 'rug pull' scam.
Aug 4, 2026
Inference-Time Policy Alignment for Fair Reinforcement Learning
Researchers propose a new inference-time method to adjust pretrained reinforcement learning agents toward fairness objectives without retraining, enabling dynamic adaptation to stakeholder preferences post-deployment.
Aug 4, 2026
Logic, Optimization, and Artificial Intelligence
A new arXiv preprint (2607.15532v1) surveys how integrating logic and optimization techniques can enhance transparency, explainability, and trustworthiness in rule-based AI systems.
Jul 20, 2026
Selective Test-Time Debiasing for CLIP via Reward Gating
Researchers propose a new method to reduce bias in vision language models.
Published Jul 2, 2026 · Analyzed Jul 5, 2026