MIT AI Risk Library: Over 700 Potential DangersAI-RESOURCESDiscover the AI Risk Library, a groundbreaking database cataloging 700+ AI risks. Understand and mitigate potential threats with expert insights.Aug 17, 2024 AI MIT
MIT's Thermometer: Calibrating Large Language ModelsAI-PAPERSNew MIT research calibrates LLMs to avoid overconfidence. Thermometer method boosts accuracy and reliability. Aug 2, 2024 AI model MIT
Mirai: An AI model to predict future breast cancerAI-PAPERSMirai predicts breast cancer risk at multiple time points, giving doctors more comprehensive information for decision-makingJul 29, 2024 AI model MIT