What is a mask in neural network?

Could you please elaborate on the concept of a "mask" in the context of neural networks? Specifically, how does it function and what role does it play in the network's overall operation? Are masks used primarily for regularization or are they employed for other purposes as well? Additionally, are there different types of masks and if so, how do they differ from each other? Lastly, could you provide an example or two to further illustrate the concept of a mask in a neural network? Thank you for your time and assistance in clarifying this aspect of neural network architectures.