English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Density functional theory (DFT) stands as a cornerstone method incomputational quantum chemistry and materials science due to its remarkableversatility and scalability. Yet, it suffers from limitations in accuracy,particularly when dealing with strongly correlated systems. To address theseshortcomings, recent work has begun to explore how machine learning can expandthe capabilities of DFT; an endeavor with many open questions and technicalchallenges. In this work, we present Grad DFT: a fully differentiable JAX-basedDFT library, enabling quick prototyping and experimentation with machinelearning-enhanced exchange-correlation energy functionals. Grad DFT employs apioneering parametrization of exchange-correlation functionals constructedusing a weighted sum of energy densities, where the weights are determinedusing neural networks. Moreover, Grad DFT encompasses a comprehensive suite ofauxiliary functions, notably featuring a just-in-time compilable and fullydifferentiable self-consistent iterative procedure. To support training andbenchmarking efforts, we additionally compile a curated dataset of experimentaldissociation energies of dimers, half of which contain transition metal atomscharacterized by strong electronic correlations. The software library is testedagainst experimental results to study the generalization capabilities of aneural functional across potential energy surfaces and atomic species, as wellas the effect of training data noise on the resulting model accuracy.

Grad DFT: a software library for machine learning enhanced density functional theory