TruthfulQA: Measuring How Models Mimic Human Falsehoods Posted on 2021/09/16 by Carl Robitaille TruthfulQA: Measuring How Models Mimic Human Falsehoods Oh great. The more complexity you include in your AI model, the better it becomes at giving you false answers. https://t.co/10ipLFitzi— Sean Carroll (@seanmcarroll) September 16, 2021