Artificial Intelligence Stack Exchange is a question and answer site for people interested in conceptual questions about life and challenges in a world where "cognitive" functions can be mimicked in purely digital environment. It only takes a minute to sign up.
Sign up to join this community
In this article, the term "learned emulator" is used.
Recently, scientists have started creating "learned emulators" using AI neural network approaches, but have not yet fully explored the advantages and potential pitfalls of these surrogates.
What is a "learned emulator"? I believe it is related to neural networks. Where can I read more?
In a typical situation, for the emulation of physical environments, you need to define all physical rules and forces. In the "learned emulators", they use some machine learning techniques to learn those rules by supervising and interacting (instead of formalizing all of them). In this case, they do not need any exact formulation of the physical environment to emulate it.
An instance of this simulation can be this article "Realistic Atomistic Structure of Amorphous Silicon from Machine-Learning-Driven Molecular Dynamics", i.e., a machine-learning-driven simulation instead of exact formulation of the environment.
