1 paper · 1 filter
Newman Cheng, Gordon Broadbent, Jason Dong +16
As AI systems become increasingly capable scientific collaborators, a fundamental challenge that arises is not only generating explanations, but calibrating confidence in them. We…