Tsung-Han Lee
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Tsung-Han Lee has a total of 43 co-authors. Here are the top co-authors by number of shared publications:
- Nicola Lanatà
Rochester Institute of Technology
8 shared publications Gabriel Kotliar
Rutgers University
6 shared publications- Yong-Xin Yao
Ames National Laboratory
4 shared publications - Vladimir Dobrosavljević
National High Magnetic Field Laboratory
3 shared publications - Eduardo Miranda
The University of Campinas
2 shared publications - Marius S. Frank
Aarhus University
2 shared publications
Latest Publications
- Open Access
- 28 July 2026
- Published by American Physical Society (APS) in PRX Intelligence
- Vol. 1 (1)
Linear Foundation Model for Quantum Embedding: Data-Driven Compression of the Ghost Gutzwiller Variational Space
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 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 solution to a deterministic ground-state eigenvalue problem in the reduced space. Within the ghost-Gutzwiller approximation, 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 simulations of strongly correlated materials at a near-DFT cost.
- 28 July 2026
- Published by American Physical Society (APS) in PRX Intelligence
- Vol. 1 (1)
Open AccessLinear Foundation Model for Quantum Embedding: Data-Driven Compression of the Ghost Gutzwiller Variational Space
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 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 solution to a deterministic ground-state eigenvalue problem in the reduced space. Within the ghost-Gutzwiller approximation, 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 simulations of strongly correlated materials at a near-DFT cost.
- 10 April 2026
- Published by American Physical Society (APS) in Physical Review B
- Vol. 113 (15)
Neural-quantum-states impurity solver for quantum embedding problems
- 10 April 2026
- Published by American Physical Society (APS) in Physical Review B
- Vol. 113 (15)
Neural-quantum-states impurity solver for quantum embedding problems
- 13 September 2024
- Published by American Physical Society (APS) in Physical Review B
- Vol. 110 (11)
Charge self-consistent density functional theory plus ghost rotationally invariant slave-boson theory for correlated materials
- 13 September 2024
- Published by American Physical Society (APS) in Physical Review B
- Vol. 110 (11)
Charge self-consistent density functional theory plus ghost rotationally invariant slave-boson theory for correlated materials
- 19 December 2023
- Published by American Physical Society (APS) in Physical Review B
- Vol. 108 (24) , 245147
Accuracy of ghost-rotationally-invariant slave-boson theory for multiorbital Hubbard models and realistic materials
We assess the accuracy of ghost-rotationally-invariant slave-boson (g-RISB) theory in multiorbital systems by applying it to both the three-orbital degenerate Hubbard model and a realistic model extracted from first-principle simulations and comparing the results to those obtained using dynamical mean-field theory (DMFT). Our findings indicate that g-RISB's accuracy can be systematically improved toward the exact DMFT limit in infinite-dimensional multiorbital models by increasing the number of ghost orbitals. This allows for a more precise description of aspects of Hund metal physics and Mott physics compared with the original RISB approach. We also demonstrate that g-RISB reliably captures the quasiparticle weights, Fermi surface, and low-energy spectral function for the realistic model compared with DMFT. Moreover, we showcase the potential of using the density matrix renormalization group method as an impurity solver within the g-RISB framework to study systems with a larger number of ghost orbitals. These results show the potential of g-RISB as a reliable tool for simulating correlated materials. The connection between the g-RISB and DMFT self-energy is also discussed.- 19 December 2023
- Published by American Physical Society (APS) in Physical Review B
- Vol. 108 (24) , 245147
Accuracy of ghost-rotationally-invariant slave-boson theory for multiorbital Hubbard models and realistic materials
We assess the accuracy of ghost-rotationally-invariant slave-boson (g-RISB) theory in multiorbital systems by applying it to both the three-orbital degenerate Hubbard model and a realistic model extracted from first-principle simulations and comparing the results to those obtained using dynamical mean-field theory (DMFT). Our findings indicate that g-RISB's accuracy can be systematically improved toward the exact DMFT limit in infinite-dimensional multiorbital models by increasing the number of ghost orbitals. This allows for a more precise description of aspects of Hund metal physics and Mott physics compared with the original RISB approach. We also demonstrate that g-RISB reliably captures the quasiparticle weights, Fermi surface, and low-energy spectral function for the realistic model compared with DMFT. Moreover, we showcase the potential of using the density matrix renormalization group method as an impurity solver within the g-RISB framework to study systems with a larger number of ghost orbitals. These results show the potential of g-RISB as a reliable tool for simulating correlated materials. The connection between the g-RISB and DMFT self-energy is also discussed.- 15 March 2023
- Published by American Physical Society (APS) in Physical Review B
- Vol. 107 (12) , L121104
Accuracy of ghost rotationally invariant slave-boson and dynamical mean field theory as a function of the impurity-model bath size
We compare the accuracy of the ghost rotationally invariant slave-boson (g-RISB) theory and dynamical mean field theory (DMFT) on the single-band Hubbard model, as a function of the number of bath sites in the embedding impurity Hamiltonian. Our benchmark calculations confirm that the accuracy of g-RISB can be systematically improved by increasing the number of bath sites, similar to DMFT. With a few bath sites, we observe that g-RISB is systematically more accurate than DMFT for the ground-state observables. On the other hand, the relative accuracy of these methods is generally comparable for the quasiparticle weight and the spectral function. As expected, we observe that g-RISB satisfies the variational principle in infinite dimensions, as the total energy decreases monotonically towards the exact value as a function of the number of bath sites, suggesting that the g-RISB wave function may approach the exact ground state in infinite dimensions. Our results suggest that the g-RISB is a promising method for first-principles simulations of strongly correlated matter, which can capture the behavior of both static and dynamical observables, at a relatively low computational cost.
- 15 March 2023
- Published by American Physical Society (APS) in Physical Review B
- Vol. 107 (12) , L121104
Accuracy of ghost rotationally invariant slave-boson and dynamical mean field theory as a function of the impurity-model bath size
We compare the accuracy of the ghost rotationally invariant slave-boson (g-RISB) theory and dynamical mean field theory (DMFT) on the single-band Hubbard model, as a function of the number of bath sites in the embedding impurity Hamiltonian. Our benchmark calculations confirm that the accuracy of g-RISB can be systematically improved by increasing the number of bath sites, similar to DMFT. With a few bath sites, we observe that g-RISB is systematically more accurate than DMFT for the ground-state observables. On the other hand, the relative accuracy of these methods is generally comparable for the quasiparticle weight and the spectral function. As expected, we observe that g-RISB satisfies the variational principle in infinite dimensions, as the total energy decreases monotonically towards the exact value as a function of the number of bath sites, suggesting that the g-RISB wave function may approach the exact ground state in infinite dimensions. Our results suggest that the g-RISB is a promising method for first-principles simulations of strongly correlated matter, which can capture the behavior of both static and dynamical observables, at a relatively low computational cost.
- 3 October 2022
- Published by American Physical Society (APS) in Physical Review B
- Vol. 106 (16) , L161102
Mott domain walls: A (strongly) non-Fermi liquid state of matter
Most Mott systems display a low-temperature phase coexistence region around the metal-insulator transition. The domain walls separating the respective phases have very recently been observed displaying unusual properties both in simulations and in experiments. First, they often cover a significant volume fraction, thus cannot be neglected. Second, they resemble neither a typical metal nor a standard insulator, displaying unfamiliar temperature dependence of (local) transport properties. Here we take a closer look at such domain wall matter by examining an appropriate unstable solution of the Hubbard model. We show that transport in this regime is dominated by the emergence of “resilient quasiparticles” displaying strong non-Fermi liquid features, reflecting the quantum-critical fluctuations in the vicinity of the Mott point.
- 3 October 2022
- Published by American Physical Society (APS) in Physical Review B
- Vol. 106 (16) , L161102
Mott domain walls: A (strongly) non-Fermi liquid state of matter
Most Mott systems display a low-temperature phase coexistence region around the metal-insulator transition. The domain walls separating the respective phases have very recently been observed displaying unusual properties both in simulations and in experiments. First, they often cover a significant volume fraction, thus cannot be neglected. Second, they resemble neither a typical metal nor a standard insulator, displaying unfamiliar temperature dependence of (local) transport properties. Here we take a closer look at such domain wall matter by examining an appropriate unstable solution of the Hubbard model. We show that transport in this regime is dominated by the emergence of “resilient quasiparticles” displaying strong non-Fermi liquid features, reflecting the quantum-critical fluctuations in the vicinity of the Mott point.
- Open Access
- 29 November 2021
- Published by American Physical Society (APS) in Physical Review X
- Vol. 11 (4) , 041040
Efficient Slave-Boson Approach for Multiorbital Two-Particle Response Functions and Superconductivity
We develop an efficient approach for computing two-particle response functions and interaction vertices for multiorbital strongly correlated systems based on the rotationally invariant slave-boson framework. The method is applied to the degenerate three-orbital Hubbard-Kanamori model for investigating the origin of the -wave orbital antisymmetric spin-triplet superconductivity in Hund’s metal regime, previously found in the dynamical mean-field theory studies. By computing the pairing interaction considering the particle-particle and the particle-hole scattering channels, we identify the mechanism leading to the pairing instability around Hund’s metal crossover arises from the particle-particle channel, which contains the local electron pair fluctuation between different particle-number sectors of the atomic Hilbert space. On the other hand, the particle-hole spin fluctuations induce the -wave pairing instability before entering Hund’s regime. Our approach paves the way for investigating the pairing mechanism in realistic correlated materials.- 29 November 2021
- Published by American Physical Society (APS) in Physical Review X
- Vol. 11 (4) , 041040
Open AccessEfficient Slave-Boson Approach for Multiorbital Two-Particle Response Functions and Superconductivity
We develop an efficient approach for computing two-particle response functions and interaction vertices for multiorbital strongly correlated systems based on the rotationally invariant slave-boson framework. The method is applied to the degenerate three-orbital Hubbard-Kanamori model for investigating the origin of the -wave orbital antisymmetric spin-triplet superconductivity in Hund’s metal regime, previously found in the dynamical mean-field theory studies. By computing the pairing interaction considering the particle-particle and the particle-hole scattering channels, we identify the mechanism leading to the pairing instability around Hund’s metal crossover arises from the particle-particle channel, which contains the local electron pair fluctuation between different particle-number sectors of the atomic Hilbert space. On the other hand, the particle-hole spin fluctuations induce the -wave pairing instability before entering Hund’s regime. Our approach paves the way for investigating the pairing mechanism in realistic correlated materials. - Open Access
- 5 August 2021
- Published by American Physical Society (APS) in Physical Review B
- Vol. 104 (8) , L081103
Quantum embedding description of the Anderson lattice model with the ghost Gutzwiller approximation
We present benchmark calculations of the Anderson lattice model based on the recently developed “ghost Gutzwiller approximation”. Our analysis shows that, in some parameters regimes, the predictions of the standard Gutzwiller approximation can be incorrect by orders of magnitude for this model. We show that this is caused by the inability of this method to describe simultaneously the Mott physics and the hybridization between correlated and itinerant degrees of freedom—whose interplay often governs the metal-insulator transition in real materials. Finally, we show that the ghost Gutzwiller approximation solves this problem, providing us with results in remarkable agreement with dynamical mean-field theory throughout the entire phase diagram, while being much less computationally demanding. We provide an analytical explanation of these findings and discuss their implications within the context of ab-initio computation of strongly correlated matter.
- 5 August 2021
- Published by American Physical Society (APS) in Physical Review B
- Vol. 104 (8) , L081103
Open AccessQuantum embedding description of the Anderson lattice model with the ghost Gutzwiller approximation
We present benchmark calculations of the Anderson lattice model based on the recently developed “ghost Gutzwiller approximation”. Our analysis shows that, in some parameters regimes, the predictions of the standard Gutzwiller approximation can be incorrect by orders of magnitude for this model. We show that this is caused by the inability of this method to describe simultaneously the Mott physics and the hybridization between correlated and itinerant degrees of freedom—whose interplay often governs the metal-insulator transition in real materials. Finally, we show that the ghost Gutzwiller approximation solves this problem, providing us with results in remarkable agreement with dynamical mean-field theory throughout the entire phase diagram, while being much less computationally demanding. We provide an analytical explanation of these findings and discuss their implications within the context of ab-initio computation of strongly correlated matter.
- Open Access
- 29 January 2021
- Published by American Physical Society (APS) in Physical Review Research
- Vol. 3 (1) , 013101
Bypassing the computational bottleneck of quantum-embedding theories for strong electron correlations with machine learning
A cardinal obstacle to performing quantum-mechanical simulations of strongly correlated matter is that, with the theoretical tools presently available, sufficiently accurate computations are often too expensive to be ever feasible. Here we design a computational framework combining quantum-embedding (QE) methods with machine learning. This allows us to bypass altogether the most computationally expensive components of QE algorithms, making their overall cost comparable to bare density functional theory. We perform benchmark calculations of a series of actinide systems, where our method accurately describes the correlation effects, reducing by orders of magnitude the computational cost. We argue that, by producing a larger-scale set of training data, it will be possible to apply our method to systems with arbitrary stoichiometries and crystal structures, paving the way to virtually infinite applications in condensed matter physics, chemistry, and materials science.
- 29 January 2021
- Published by American Physical Society (APS) in Physical Review Research
- Vol. 3 (1) , 013101
Open AccessBypassing the computational bottleneck of quantum-embedding theories for strong electron correlations with machine learning
A cardinal obstacle to performing quantum-mechanical simulations of strongly correlated matter is that, with the theoretical tools presently available, sufficiently accurate computations are often too expensive to be ever feasible. Here we design a computational framework combining quantum-embedding (QE) methods with machine learning. This allows us to bypass altogether the most computationally expensive components of QE algorithms, making their overall cost comparable to bare density functional theory. We perform benchmark calculations of a series of actinide systems, where our method accurately describes the correlation effects, reducing by orders of magnitude the computational cost. We argue that, by producing a larger-scale set of training data, it will be possible to apply our method to systems with arbitrary stoichiometries and crystal structures, paving the way to virtually infinite applications in condensed matter physics, chemistry, and materials science.
- 2 June 2020
- Published by American Physical Society (APS) in Physical Review B
- Vol. 101 (23) , 235112
Two-dimensional disordered Mott metal-insulator transition
We studied several aspects of the Mott metal-insulator transition in the disordered case. The model on which we based our analysis is the disordered Hubbard model, which is the simplest model capable of capturing the Mott metal-insulator transition. We investigated this model through statistical dynamical mean-field theory (statDMFT). This theory is a natural extension of dynamical mean-field theory (DMFT), which has been used with relative success in the past several years with the purpose of describing the Mott transition in the clean case. As is the case for the latter theory, statDMFT incorporates the electronic correlation effects only in their local manifestations. Disorder, on the other hand, is treated in such a way as to incorporate Anderson localization effects. With this technique, we analyzed the disordered two-dimensional Mott transition, using the quantum Monte Carlo algorithm to solve the associated single-impurity problems. We found spinodal lines at which the metal and insulator ceased to be metastable. We also studied spatial fluctuations of local quantities, such as self-energy and local Green's function, and showed the appearance of metallic regions within the insulator and vice versa. We carried out an analysis of finite-size effects and showed that, in agreement with the theorems of Imry and Ma [Y. Imry and S. K. Ma, Phys. Rev. Lett. 35, 1399 (1975).], the first-order transition is smeared in the thermodynamic limit. We analyzed transport properties by means of a mapping to a random classical resistor network and calculated both the average current and its distribution across the metal-insulator transition.
- 2 June 2020
- Published by American Physical Society (APS) in Physical Review B
- Vol. 101 (23) , 235112
Two-dimensional disordered Mott metal-insulator transition
We studied several aspects of the Mott metal-insulator transition in the disordered case. The model on which we based our analysis is the disordered Hubbard model, which is the simplest model capable of capturing the Mott metal-insulator transition. We investigated this model through statistical dynamical mean-field theory (statDMFT). This theory is a natural extension of dynamical mean-field theory (DMFT), which has been used with relative success in the past several years with the purpose of describing the Mott transition in the clean case. As is the case for the latter theory, statDMFT incorporates the electronic correlation effects only in their local manifestations. Disorder, on the other hand, is treated in such a way as to incorporate Anderson localization effects. With this technique, we analyzed the disordered two-dimensional Mott transition, using the quantum Monte Carlo algorithm to solve the associated single-impurity problems. We found spinodal lines at which the metal and insulator ceased to be metastable. We also studied spatial fluctuations of local quantities, such as self-energy and local Green's function, and showed the appearance of metallic regions within the insulator and vice versa. We carried out an analysis of finite-size effects and showed that, in agreement with the theorems of Imry and Ma [Y. Imry and S. K. Ma, Phys. Rev. Lett. 35, 1399 (1975).], the first-order transition is smeared in the thermodynamic limit. We analyzed transport properties by means of a mapping to a random classical resistor network and calculated both the average current and its distribution across the metal-insulator transition.