Molecular Electronic-Structure Theory

Our current activities span much of the spectrum of electronic-structure theory, all the way from developing methods with the capabilities to simulate modest-sized systems at near-exact accuracy to developing widely applicable methods operating at a much more limited level of accuracy.

The need for calibration and advancements on leading approximations to formally exact theory is highly relevant; possibly more so today than throughout the past two or three decades. This is particularly with the advent of novel forms of computation in mind. In our research activities, we keep this central. Simultaneously, we work under the belief that simulations which lend themselves to established methods in quantum chemistry can benefit from being viewed through more localized and alternative perspectives. This shift has the potential to refine and advance our understanding of how best to approximate the electron correlation problem.

Finally, we break down molecular simulations into contributions from individual atoms or functional groups which allows for a more precise resolution of the training data. This can significantly improve the accuracy of machine-learning models in chemistry, which is an area we are increasingly interested in contributing to.