
To help people better anticipate a self-driving car’s mistakes, a team led by @MITAeroAstro Professor Julie Shah developed a new method that provides clear explanations of the underlying model’s decisions, as described in a @Nature paper. news.mit.edu/2026/system-hel…



