arXiv subjects

Soohaeng Yoo Willow

Publications and source records attributed to Soohaeng Yoo Willow.

13 recordsLinked to original sources

arXiv2026

GW and Bethe-Salpeter Theory for Molecular Polaritons, Quasiparticles, and Excitons

The electron self-energy is central to quasiparticle theory, yet how an optical cavity enters it remains unclear. We address this question for a molecule in a single-mode cavity using the dipole-gauge Pauli-Fierz Hamiltonian and a coherent-state QED Hartree-Fock reference. The cavity enters through three channels: the static dipole self-energy (DSE) shift of reference orbital energies, direct DSE augmentation of the screened interaction, and the polariton pole carrying the bilinear electron-photon coupling. We benchmark QED-$GW$ ionization potentials (IPs) and electron affinities (EAs) against a cavity $Δ$-method ladder from QED-HF to correlated wave-function methods, whose cavity-induced shifts agree within 1 meV where directly comparable. For closed-shell molecules with unbound anions, $GW$ systematically overestimates cavity-induced IP redshifts, whereas EA shifts are reproduced nearly quantitatively, although this does not imply comparable accuracy for absolute EAs. For ionic molecules with bound anions, this ordering reverses, consistent with published QED coupled-cluster results. Coupling and detuning scans show that the error is predominantly quadratic in $λ$ and DSE-driven rather than resonant. The spectral function develops a polariton-replica photoemission sideband with weight scaling as $λ^2$. In the static screened interaction used in the Bethe-Salpeter equation, bare-photon exchange cancels the matching DSE contribution to the direct interaction, while exchange and polariton-screening corrections remain. Their net effect on the lowest excitation is appreciable only for ammonia in the molecules studied. Exciton-binding energies involving unbound anions are strongly basis-dependent and should therefore be viewed as diagnostics of electron-hole interactions rather than basis-converged molecular quantities.

cond-mat.mtrl-sci↗

arXiv2026

Accelerated Discovery of Nitrogen-Coordinated Dual-Atom Hydrogen Evolution Reaction Electrocatalysts via Machine Learning Potentials

The hydrogen evolution reaction (HER) is central to sustainable hydrogen production, and nitrogen coordinated dual atom catalysts (DACs) offer a promising route to noble metal activity at low cost. Yet their vast compositional and coordination design space remains underexplored, as density functional theory (DFT) screening at scale is prohibitive. Here, we map the HER landscape of graphene supported TM2@Nx-Gr DACs, screening 23 transition metals across 20 nitrogen coordination motifs using a machine learning potential (MLP) benchmarked against DFT. Intermediate coordination (2N to 4N) consistently yields near-optimal ΔGH*, with Ti2@2Na, Mn2@2Na, Fe2@2Na, Cu2@2Na, Rh2@2Na, Zr2@2Na, Zr2@2Nb, Zr2@2Nc, Nb2@2Nc, Zr2@2Nd, Mn2@2Ne, Mn2@2Nf, Ti2@3Na, Au2@3Na, Fe2@3Na, Pd2@3Nb, Rh2@3Nc, Rh2@3Nd, Au2@3Nd, V2@4Na, Ti2@4Nb, Pd2@4Nb, Ti2@4Nc, Cr2@4Nd, Ni2@4Nd, Cu2@4Nd emerging as standout, synthesizable candidates, most exhibiting metallic or narrow gap (<0.25 eV) character. The MLP reaches near-DFT accuracy, with a mean absolute error of 80 meV for Gibbs binding free energies at orders of magnitude lower computational cost, establishing MLP driven screening as a practical engine for next-generation catalyst discovery.

cond-mat.mtrl-sci↗

arXiv2026

Bayesian E(3)-Equivariant Interatomic Potential with Iterative Restratification of Many-body Message Passing

Machine learning potentials (MLPs) have become essential for large-scale atomistic simulations, enabling ab initio-level accuracy with computational efficiency. However, current MLPs struggle with uncertainty quantification, limiting their reliability for active learning, calibration, and out-of-distribution (OOD) detection. We address these challenges by developing Bayesian E(3) equivariant MLPs with iterative restratification of many-body message passing. Our approach introduces the joint energy-force negative log-likelihood (NLL$_\text{JEF}$) loss function, which explicitly models uncertainty in both energies and interatomic forces, yielding substantially improved accuracy compared to conventional NLL losses. We systematically benchmark multiple Bayesian approaches, including deep ensembles with mean-variance estimation, stochastic weight averaging Gaussian, improved variational online Newton, and Laplace approximation by evaluating their performance on uncertainty prediction, OOD detection, calibration, and active learning tasks. We further demonstrate that NLL$_\text{JEF}$ facilitates efficient active learning by quantifying energy and force uncertainties. Using Bayesian active learning by disagreement (BALD), our framework outperforms random sampling and energy-uncertainty-based sampling. Our results demonstrate that Bayesian MLPs achieve competitive accuracy with state-of-the-art models while enabling uncertainty-guided active learning, OOD detection, and energy/forces calibration. This work establishes Bayesian equivariant neural networks as a powerful framework for developing uncertainty-aware MLPs for atomistic simulations at scale.

cs.LG↗

arXiv2026

Stochastic Loop Corrections to Belief Propagation for Tensor Network Contraction

Tensor network contraction is a fundamental computational challenge underlying quantum many-body physics, statistical mechanics, and machine learning. Belief propagation (BP) provides an efficient approximate solution, but introduces systematic errors on graphs with loops. Here, we introduce a hybrid method that achieves accurate results by stochastically sampling loop corrections to BP and showcase our method by applying it to the two-dimensional ferromagnetic Ising model. For any pairwise Markov random field with symmetric edge potentials, our approach exploits an exact factorization of the partition function into the BP contribution and a loop correction factor summing over all valid loop configurations, weighted by edge weights derived directly from the potentials. We sample this sum using Markov chain Monte Carlo with moves that preserve the loop constraint, combined with umbrella sampling to ensure efficient exploration across all correlation strengths. Our stochastic approach provides unbiased estimates with controllable statistical error in any parameter regime.

cond-mat.str-el↗

arXiv2025

Hybrid Quantum--Classical Machine Learning Potential with Variational Quantum Circuits

Quantum algorithms for simulating large and complex molecular systems are still in their infancy, and surpassing state-of-the-art classical techniques remains an ever-receding goal post. A promising avenue of inquiry in the meanwhile is to seek practical advantages through hybrid quantum-classical algorithms, which combine conventional neural networks with variational quantum circuits (VQCs) running on today's noisy intermediate-scale quantum (NISQ) hardware. Such hybrids are well suited to NISQ hardware. The classical processor performs the bulk of the computation, while the quantum processor executes targeted sub-tasks that supply additional non-linearity and expressivity. Here, we benchmark a purely classical E(3)-equivariant message-passing machine learning potential (MLP) against a hybrid quantum-classical MLP for predicting density functional theory (DFT) properties of liquid silicon. In our hybrid architecture, every readout in the message-passing layers is replaced by a VQC. Molecular dynamics simulations driven by the HQC-MLP reveal that an accurate reproduction of high-temperature structural and thermodynamic properties is achieved with VQCs. These findings demonstrate a concrete scenario in which NISQ-compatible HQC algorithm could deliver a measurable benefit over the best available classical alternative, suggesting a viable pathway toward near-term quantum advantage in materials modeling.

quant-ph↗

arXiv2025

Machine Learning Nonadiabatic Dynamics: Eliminating Phase Freedom of Nonadiabatic Couplings with the State-Intraction State-Averaged Spin-Restricted Ensemble-Referenced Kohn-Sham Approach

Excited-state molecular dynamics (ESMD) simulations near conical intersections (CIs) pose significant challenges when using machine learning potentials (MLPs). Although MLPs have gained recognition for their integration into mixed quantum-classical (MQC) methods, such as trajectory surface hopping (TSH), and their capacity to model correlated electron-nuclear dynamics efficiently, difficulties persist in managing nonadiabatic dynamics. Specifically, singularities at CIs and double-valued coupling elements result in discontinuities that disrupt the smoothness of predictive functions. Partial solutions have been provided by learning diabatic Hamiltonians with phaseless loss functions to these challenges. However, a definitive method for addressing the discontinuities caused by CIs and double-valued coupling elements has yet to be developed. Here, we introduce the phaseless coupling term, $Δ^2$, derived from the square of the off-diagonal elements of the diabatic Hamiltonian in the state-interaction state-averaged spin-restricted ensemble-referenced Kohn-Sham (SI-SA-REKS, briefly SSR)(2,2) formalism. This approach improves the stability and accuracy of the MLP model by addressing the issues arising from CI singularities and double-valued coupling functions. We apply this method to the penta-2,4-dieniminium cation (PSB3), demonstrating its effectiveness in improving MLP training for ML-based nonadiabatic dynamics. Our results show that the $Δ^2$ based ML-ESMD method can reproduce ab initio ESMD simulations, underscoring its potential and efficiency for broader applications, particularly in large-scale and long-timescale ESMD simulations.

physics.chem-ph↗

arXiv2024

Active Sparse Bayesian Committee Machine Potential for Isothermal-Isobaric Molecular Dynamics Simulations

Recent advancements in machine learning potentials (MLPs) have significantly impacted the fields of chemistry, physics, and biology by enabling large-scale first-principles simulations. Among different machine learning approaches, kernel-based MLPs distinguish themselves through their ability to handle small datasets, quantify uncertainties, and minimize over-fitting. Nevertheless, their extensive computational requirements present considerable challenges. To alleviate these, sparsification methods have been developed, aiming to reduce computational scaling without compromising accuracy. In the context of isothermal and isobaric ML molecular dynamics (MD) simulations, achieving precise pressure estimation is crucial for reproducing reliable system behavior under constant pressure. Despite progress, sparse kernel MLPs struggle with precise pressure prediction. Here, we introduce a virial kernel function that significantly enhances pressure estimation accuracy of MLPs. Additionally, we propose the active sparse Bayesian committee machine (BCM) potential, an on-the-fly MLP architecture that aggregates local sparse Gaussian process regression (SGPR) MLPs. The sparse BCM potential overcomes the steep computational scaling with the kernel size, and a predefined restriction on the size of kernel allows for a fast and efficient on-the-fly training. Our advancements facilitate accurate and computationally efficient machine learning-enhanced MD (MLMD) simulations across diverse systems, including ice-liquid coexisting phases, \ce{Li10Ge(PS6)2} lithium solid electrolyte, and high-pressure liquid boron nitride.

cond-mat.soft↗

arXiv2024

A Bayesian Committee Machine Potential for Oxygen-containing Organic Compounds

Understanding the pivotal role of oxygen-containing organic compounds in serving as an energy source for living organisms and contributing to protein formation is crucial in the field of biochemistry. This study addresses the challenge of comprehending protein-protein interactions (PPI) and developing predicitive models for proteins and organic compounds, with a specific focus on quantifying their binding affinity. Here, we introduce the active Bayesian Committee Machine (BCM) potential, specifically designed to predict oxygen-containing organic compounds within eight groups of CHO. The BCM potential adopts a committee-based approach to tackle scalability issues associated with kernel regressors, particularly when dealing with large datasets. Its adaptable structure allows for efficient and cost-effective expansion, maintaing both transferability and scalability. Through systematic benchmarking, we position the sparse BCM potential as a promising contender in the pursuit of a universal machine learning potential.

cond-mat.mtrl-sci↗

arXiv2024

A Bayesian Committee Machine Potential for Organic Nitrogen Compounds

Large-scale computer simulations of chemical atoms are used in a wide range of applications, including batteries, drugs, and more. However, there is a problem with efficiency as it takes a long time due to the large amount of calculation. To solve these problems, machine learning interatomic potential (ML-IAP) technology is attracting attention as an alternative. ML-IAP not only has high accuracy by faithfully expressing the density functional theory (DFT), but also has the advantage of low computational cost. However, there is a problem that the potential energy changes significantly depending on the environment of each atom, and expansion to a wide range of compounds within a single model is still difficult to build in the case of a kernel-based model. To solve this problem, we would like to develop a universal ML-IAP using this active Bayesian Committee Machine (BCM) potential methodology for carbon-nitrogen-hydrogen (CNH) with various compositions. ML models are trained and generated through first-principles calculations and molecular dynamics simulations for molecules with only CNH. Using long amine structures to test an ML model trained only with short chains, the results show excellent consistency with DFT calculations. Consequently, machine learning-based models for organic molecules not only demonstrate the ability to accurately describe various physical properties but also hold promise for investigating a broad spectrum of diverse materials systems.

cond-mat.mtrl-sci↗

arXiv2024

A Sparse Bayesian Committee Machine Potential for Hydrocarbons

Accurate and scalable universal interatomic potentials are key for understanding material properties at the atomic level, a task often hindered by the steep computational scaling. Although recent developments of machine learning potential has made significant progress, the flexibility and expansion to a wide range of compounds within a single model seems still challenging to build in particular for the kernel-based models. Here, we introduce the Bayesian Committee Machine (BCM) potential to handle a wide array of hydrocarbons in various phases (gas, clusters, liquid, and solid). The BCM potential leverages the committee approach to bypass the poor scaling of kernel regressors when dealing with large datasets. Its committee-based structure allows for easy and cost-effective expansion, maintaining both transferrability and scalability. Demonstrating its robustness, the BCM accurately models challenging examples like the Diels-Alder reaction, structural strains, and pi-pi interactions. Our systematic benchmarking positions the sparse BCM potential as a promising candidate in the quest for a universal ab initio machine learning potential.

cond-mat.mtrl-sci↗

arXiv2023

Learned Mappings for Targeted Free Energy Perturbation between Peptide Conformations

Targeted free energy perturbation uses an invertible mapping to promote configuration space overlap and the convergence of free energy estimates. However, developing suitable mappings can be challenging. Wirnsberger et al. (2020) demonstrated the use of machine learning to train deep neural networks that map between Boltzmann distributions for different thermodynamic states. Here, we adapt their approach to free energy differences of a flexible bonded molecule, deca-alanine, with harmonic biases with different spring centers. When the neural network is trained until ``early stopping'' - when the loss value of the test set increases - we calculate accurate free energy differences between thermodynamic states with spring centers separated by 1 Å and sometimes 2 Å. For more distant thermodynamic states, the mapping does not produce structures representative of the target state and the method does not reproduce reference calculations.

cond-mat.stat-mech↗

arXiv2020

On the Polarization of Ligands by Proteins

Although ligand-binding sites in many proteins contain a high number density of charged side chains that can polarize small organic molecules and influence binding, the magnitude of this effect has not been studied in many systems. Here, we use a quantum mechanics/molecular mechanics (QM/MM) approach in which the ligand is the QM region to compute the ligand polarization energy of 286 protein-ligand complexes from the PDBBind Core Set (release 2016). We observe that the ligand polarization energy is linearly correlated with the magnitude of the electric field acting on the ligand, the magnitude of the induced dipole moment, and the classical polarization energy. The influence of protein and cation charges on the ligand polarization diminishes with the distance and is below 2 kcal/mol at 9 $\unicode{x212B}$ and 1 kcal/mol at 12 $\unicode{x212B}$. Considering both polarization and solvation appears essential to computing negative binding energies in some crystallographic complexes. Solvation, but not polarization, is essential for achieving moderate correlation with experimental binding free energies.

physics.chem-ph↗

arXiv2014

The freezing tendency towards 4-coordinated amorphous network causes increase in heat capacity of supercooled Stillinger-Weber silicon

The supercooled liquid silicon, modeled by Stillinger-Weber potential, shows anomalous increase in heat capacity $C_p$, with a maximum $C_p$ value close to 1060 K at zero pressure. We study equilibration and relaxation of the supercooled SW Si, in the temperature range of 1060 K--1070 K at zero pressure. We find that as the relaxation of the metastable supercooled liquid phase initiates, a straight line region (SLR) is formed in cumulative potential energy distributions. The configurational temperature corresponding to the SLR is close to 1060 K, which was earlier identified as the freezing temperature of 4-coordinated amorphous network. The SLR is found to be tangential to the distribution of the metastable liquid phase and thus influences the broadness of the distribution. As the bath temperature is reduced from 1070 K to 1060 K, the effective temperature approaches the bath temperature which results in broadening of the metastable phase distribution. This, in turn, causes an increase in overall fluctuations of potential energy and hence an increase of heat capacity. We also find that during initial stages of relaxation, 4-coordinated atoms form 6-membered rings with a chair--like structure and other structural units that indicate crystallization. Simultaneously a strong correlation is established between the number of chair-shaped 6-membered rings and the number of 4-coordinated atoms in the system. This shows that all properties related to 4-coordinated particles are highly correlated as the SLR is formed in potential energy distributions and this can be interpreted as a consequence of `freezing' of amorphous network formed by 4-coordinated particles.

cond-mat.stat-mech↗