Liquid network cells' Neural Circuit Policy (NCP) learns what it is actually supposed to do; can perform well in conditions it has never experienced

https://news.mit.edu/2021/cause-effect-neural-networks-1014

"These neural networks know what they’re doing... can learn the cause-and-effect basis of a navigation task during training... could improve the reliability and trustworthiness of machine learning agents that are performing high-stakes tasks, like driving an autonomous vehicle on a busy highway"

Earlier referenced article: “Liquid” machine-learning system adapts to changing conditions

https://news.mit.edu/2021/machine-learning-adapts-0128

"new type of neural network could aid decision making in autonomous driving and medical diagnosis"