https://derekwang2002.github.io/ai-intelligence/en/projects/fairstress/ Material changes in AI models, agents, engineering and infrastructure, with primary sources and recommendations. en Mon, 02 Nov 2026 16:56:00 GMT https://derekwang2002.github.io/ai-intelligence/en/events/ev-20261021-07/ ev-20261021-07-published Wed, 21 Oct 2026 15:58:23 GMT New TRIAL <p>Ai2 released FairStress, an open benchmark that tests whether matched persuasive prefixes change LLM decisions differently across demographic groups. It expands 300 scenarios across 29 high-stakes domains into more than 13 million cases, with a 50K-item core subset, public CC-BY-4.0 data and code. The benchmark uses an equality-centered definition of fairness and does not cover every equity framework.</p><p><strong>Why read: </strong>FairStress exposes unstable decision criteria even when final answers match, making it useful for predeployment stress tests.</p><p>Recommendation: TRIAL</p><p>Sources: <a href="https://huggingface.co/blog/allenai/fairstress">FairStress launch</a> · <a href="https://github.com/pardissz/fairstress">FairStress code</a> · <a href="https://huggingface.co/datasets/allenai/fairstress">FairStress dataset</a></p>