Accurate and scalable exchange-correlation with deep learning

2025-06-17

AI Summary

How to construct exchange-correlation (XC) functionals that simultaneously achieve chemical accuracy (<1 kcal/mol) and computational scalability? This work introduces Skala, a deep learning–driven XC functional that learns electron correlation directly from high-fidelity CCSD(T) wavefunction data—without hand-crafted features. Methodologically, Skala integrates physical constraints (e.g., size consistency, exact exchange asymptotics) with scalable neural network architectures, ensuring full compatibility with standard density functional theory (DFT) implementations. It is the first deep learning–based XC functional to attain chemical accuracy on small-molecule atomization energies. When trained on extended datasets, Skala matches state-of-the-art hybrid functionals across mainstream main-group chemistry tasks—including reaction energies and conformational energies—while incurring only the computational cost of semilocal DFT. This work establishes a new paradigm for designing general-purpose, highly accurate, and computationally efficient DFT functionals.

Technology Category

• Machine Learning: Scalability of ML Systems
• Data Mining & Knowledge Management: Scalability, Parallel & Distributed Systems
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Density Functional Theory (DFT) is the most widely used electronic structure method for predicting the properties of molecules and materials. Although DFT is, in principle, an exact reformulation of the Schr""odinger equation, practical applications rely on approximations to the unknown exchange-correlation (XC) functional. Most existing XC functionals are constructed using a limited set of increasingly complex, hand-crafted features that improve accuracy at the expense of computational efficiency. Yet, no current approximation achieves the accuracy and generality for predictive modeling of laboratory experiments at chemical accuracy -- typically defined as errors below 1 kcal/mol. In this work, we present Skala, a modern deep learning-based XC functional that bypasses expensive hand-designed features by learning representations directly from data. Skala achieves chemical accuracy for atomization energies of small molecules while retaining the computational efficiency typical of semi-local DFT. This performance is enabled by training on an unprecedented volume of high-accuracy reference data generated using computationally intensive wavefunction-based methods. Notably, Skala systematically improves with additional training data covering diverse chemistry. By incorporating a modest amount of additional high-accuracy data tailored to chemistry beyond atomization energies, Skala achieves accuracy competitive with the best-performing hybrid functionals across general main group chemistry, at the cost of semi-local DFT. As the training dataset continues to expand, Skala is poised to further enhance the predictive power of first-principles simulations.

Problem

• Developing accurate deep learning-based exchange-correlation functionals for DFT
• Achieving chemical accuracy in molecular properties with computational efficiency
• Enhancing predictive power of DFT through scalable data-driven training

Innovation

• Deep learning-based XC functional Skala
• Learns representations directly from data
• Trains on high-accuracy reference data