Linear Foundation Model for Quantum Embedding: Data-Driven Compression of the Ghost Gutzwiller Variational Space
Samuele Giuli, Hasanat Hasan, Benedikt Kloss, Marius S. Frank, Tsung-Han Lee, Olivier Gingras, Yong-Xin Yao, Nicola Lanatà
TL;DR
This work addresses the computational bottleneck in quantum embedding methods by introducing a PCA-based linear foundation model that learns a compact variational subspace for the embedding Hamiltonian. By partitioning EH parameter space into charts and constructing an optimal subspace per chart, the EH is reduced to a small deterministic eigenproblem, guaranteeing physical constraints such as N-representability. The authors validate the approach in ghost-GA for a three-orbital Kanamori model and across Pu phases, demonstrating transferability across lattice geometries and orders-of-magnitude speedups with negligible energy error. An active-learning extension further reduces training needs, enabling robust high-throughput ab initio simulations of strongly correlated materials at near-DFT cost, and the framework is compatible with advanced solvers via the method of snapshots and potential neural-network quantum state hybrids.
Abstract
Simulations of quantum matter rely mainly on Kohn-Sham density functional theory (DFT), which often fails for strongly correlated systems. Quantum embedding (QE) theories address this limitation by mapping the system onto an auxiliary embedding Hamiltonian (EH) describing fragment-environment interactions, but the EH is typically large and its iterative solution is the primary computational bottleneck. We introduce a linear foundation model for QE that utilizes principal component analysis (PCA) to compress the space of quantum states needed to solve the EH within a small variational subspace. Using a data-driven active-learning scheme, we learn this subspace from EH ground states and reduce each embedding solve to a deterministic ground-state eigenvalue problem in the reduced space. Within the ghost Gutzwiller approximation (ghost-GA), we show for a three-orbital Hubbard model that a variational space learned on a Bethe lattice is transferable to square and cubic lattices without additional training, while substantially reducing the cost of the EH step. We further validate the approach on plutonium, where a single variational space reproduces the energetics of all six crystalline phases while reducing the cost of the EH solution by orders of magnitude. This provides a practical route to overcome the main computational bottleneck of QE frameworks, paving the way for high-throughput ab initio simulations of strongly correlated materials at a near-DFT cost.