important work being done by AI safety researcher @SharonYixuanLi Her out-of-distribution (OOD) detection feature is helping address risks not present in initial training data – a important piece of the puzzle as these models advance https://
technologyreview.com/2023/09/12/107
8351/sharon-li-ai-innovation/
…

→ View original post on X — @mustafasuleyman,