Searcharxiv⌕ Search

arXiv subjects

Honghui Shang

Publications and source records attributed to Honghui Shang.

At least 19 recordsLinked to original sources

Multi-reference variational states resolve stripe selection and local $d$-wave pairing in the Hubbard model

Whether stripes compete with $d$-wave superconductivity or promote pairing through fluctuations remains unsettled in the two-dimensional Hubbard model. Using multi-reference neural variational auxiliary-field quantum Monte Carlo (MR-nVAFQMC) at $U/t=8$, we superpose unpaired Slater determinants across stripe periods and domain-wall positions, reaching variational energies competitive with leading benchmarks on periodic lattices up to $32\times32$ sites. For cuprate next-nearest-neighbor hopping ($t'/t=-0.2$), diagonal hopping preserves electron-doped antiferromagnetism while kinetically frustrating hole motion into Yamada stripes with a dome of short-range connected $d$-wave pairing. At $1/8$ hole doping, hole delocalization across antiphase spin boundaries outweighs on-site repulsion, splitting the filled stripes of the pure model into half-filled walls with strong local pairing. Coherently delocalizing shifted walls strongly suppresses static charge modulation yet leaves short-range pairing unchanged, showing that local pairing is intrinsic to half-filled walls rather than driven by stripe fluctuations. Across intervening antiferromagnetic domains, however, pairing correlations become unresolved at one stripe period and beyond. Both parameter transfer to larger lattices without reoptimization and deep benchmark checkpoints retain stripe spin order without resolving long-range pairing, within uncertainties allowing weak superconductivity. Half-filled antiphase walls thus generate local $d$-wave pairs, whereas suppressing static stripe order alone yields no resolved coherence across stripes in the states studied.

cond-mat.mtrl-sci↗

Fully GPU-accelerated semistochastic Epstein-Nesbet perturbation theory

Second-order Epstein-Nesbet perturbation theory (EN-PT2) is an important tool for improving selected configuration interaction (SCI) energies, but the large external determinant space creates substantial memory and communication costs on GPU clusters. We develop a fully GPU-accelerated implementation of distributed semistochastic EN-PT2, fusing excitation generation, screening, and PT2 weighting in a single kernel traversal. Fixed-size random-phase CountSketch arrays limit memory use and replace large external determinant exchange with bounded collective summation. The implementation on one A100 completes EN-PT2 calculations 8.21-37.53 times faster than Arrow on 64 CPU cores at the same requested statistical standard error, and fixed-workload strong-scaling tests achieve 93.5% parallel efficiency on 16 GPUs across two nodes. Applications to an iron-oxo complex in a (60e, 187o) active space, iron-porphyrin references containing 200 million determinants per state, and neural-network quantum states establish the method's efficiency for perturbative corrections to large reference wave functions in the GPU era.

physics.comp-ph↗

Exact verification of Givens rotation circuits for efficient quantum phase estimation of iron-sulfur clusters

Givens rotations with angles stored in quantum registers account for much of the cost of published quantum phase estimation algorithms based on double-factorized tensor hypercontraction and spectral amplification. These rotations create superpositions and cannot be verified directly by the basis-state evaluators used to optimize classical reversible circuits. We give an exact verification method that separates out a fixed Clifford change of basis and tracks basis states and phases in the remaining circuit. Integer conditions establish correctness of the full coherent action, including measurement-based uncomputation and restoration of the phase-gradient resource state. Under our interpretation of Low et al.'s angle grid and rotation convention, simplifying the adders lowers the rotation cost relative to the published accounting; fusing the two additions into one masked addition gives an additional saving. We verify both implementations exhaustively at every precision used in the published walks and prove the fused circuit correct for all widths. For benchmarks including FeMoco and other iron-sulfur clusters, replacing only the rotation term lowers total phase-estimation Toffoli counts by 9.5-14.3% under the published error budget, with the walk operator unchanged.

quant-ph↗

Foundation Neural-Network Quantum States for Molecular Potential Energy Surfaces in Second Quantization

Second-quantized neural-network quantum states have achieved accurate molecular energies, but extending them across molecular geometries requires a shared representation of the geometry-dependent wavefunction coefficients. We introduce geometry-conditioned foundation neural-network quantum states for molecular electronic structure in second quantization. A single autoregressive model learns a family of ground states from sparse anchor geometries and provides wavefunctions at untrained geometries without further optimization. Orbital alignment matches orbital identities and transports their phases, establishing an aligned orbital basis across geometries. Frozen energies reach chemical accuracy at every untrained query geometry for N$_2$, CO, and H$_4$. On additional molecular paths, the energy-trained wavefunctions yield dipoles, quadrupoles, and natural occupations without property labels. Across three paired N$_2$ training seeds, orbital alignment lowers the mean absolute energy error over all untrained query geometries from 34-37 mHa to 0.049-0.085 mHa. At approximately 1 mHa mean absolute error, frozen evaluation reduces the per-geometry cost by $986\times$ relative to independent optimization, yielding an estimated $25.8\times$ end-to-end GPU-cost reduction on a 161-point N$_2$ grid.

physics.chem-ph↗

Scaling Neural Network Quantum States for Ab Initio Quantum Chemistry

Neural-network quantum states (NNQSs) can represent many-electron wave functions without explicitly enumerating the determinant space, but their accuracy depends jointly on model size and variational-optimization effort. Here we characterize this dependence for a physics-conditioned autoregressive NNQS trained separately on two six-molecule source benchmarks. Across eight model sizes and five optimization milestones, we find that model size and optimization steps jointly shape the energy error. The capacity advantage of larger models becomes more apparent with sufficient optimization, while the returns from additional optimization vary with model size. We capture this coupling using an interaction scaling law and quantify the cumulative compute of each evaluated configuration. The resulting error-compute Pareto frontiers provide a practical decision rule for jointly selecting model size and optimization steps under a given compute budget within the evaluated range. Furthermore, we find that this beneficial scaling trend persists during fine-tuning on held-out N$_2$. Pretrained models show decreasing error with increasing model size, with a steeper reduction following pretraining on the Hard benchmark. Together, these results place autoregressive neural quantum states within the broader landscape of empirical neural scaling and open a quantitative route toward the systematic scaling of neural quantum solvers for ab initio quantum chemistry.

physics.chem-ph↗

Transformer refined quantum sampling for strongly correlated electronic structure

Although quantum computing offers a promising solution for strongly correlated system simulation, existing algorithms face significant bottlenecks on current noisy intermediate-scale quantum (NISQ) devices. Here, we introduce QiankunNet-QSCI, a hybrid quantum-classical framework that addresses this challenge by combining efficient quantum-sampling with a transformer neural network. An efficient unitary selected configuration Interaction (USCI) ansatz especially designed for quantum sampling is proposed to identify the most chemically significant electronic configurations on the Zuchongzhi 3.1 quantum processor. Subsequently, the transformer model QiankunNet learns from these sparse yet critical quantum data to infer and reconstruct the complete electronic wavefunction with high fidelity. Simulation of the challenging 40-qubit [2Fe-2S] ferredoxin active center achieves chemical accuracy. Simulation of the nitrogenase P-cluster in a 114-electron 73-orbital active space also reaches 12 milli-Hartree-level agreement with the best density matrix renormalization group (DMRG) result. QiankunNet-QSCI thus offers a practical route to accurate quantum-assisted electronic structure calculations on current devices.

quant-ph↗

Implementation of the hybrid exchange-correlation functionals in the SIESTA code

We present an efficient and accurate implementation of hybrid exchange-correlation (XC) functionals in the SIESTA code, enabling large-scale simulations based on Hartree-Fock-type exact exchange combined with strictly localized numerical atomic orbitals (NAOs). Our approach exploits a fitted representation of the NAOs in terms of Gaussian-type orbitals (GTOs), which allows for the analytical evaluation of four-center electron repulsion integrals (ERIs) via the LIBINT library. This framework is seamlessly integrated with SIESTA's real-space grid and sparse-matrix infrastructure, and is combined with multiple screening techniques to control the computational complexity. We also introduce a fully analytical formulation of hybrid-functional forces and a dynamic parallel distribution scheme that ensures excellent scalability. We validate our implementation through benchmark calculations on a broad set of systems (including semiconductors, insulators, and two-dimensional materials) and demonstrate that the HSE06 functional significantly improves the prediction of band gaps compared to PBE, in close agreement with G0W0 and experimental data. We analyze in detail the trade-offs between accuracy and computational efficiency as a function of the number of Gaussians, basis set range, and integral screening thresholds. Our results confirm that hybrid functional calculations in SIESTA are now feasible for large extended systems, making accurate first-principles predictions of electronic and structural properties accessible at scale.

cond-mat.mtrl-sci↗

A Fully GPU-Accelerated Framework for High-Performance Configuration Interaction Selection with Neural Network Quantum States

AI-driven methods have demonstrated considerable success in tackling the central challenge of accurately solving the Schrödinger equation for complex many-body systems. Among neural network quantum state (NNQS) approaches, the NNQS-SCI (Selected Configuration Interaction) method stands out as a state-of-the-art technique, recognized for its high accuracy and scalability. However, its application to larger systems is severely constrained by a hybrid CPU-GPU architecture. Specifically, centralized CPU-based global de-duplication creates a severe scalability barrier due to communication bottlenecks, while host-resident coupled-configuration generation induces prohibitive computational overheads. We introduce QiankunNet-cuSCI, a fully GPU-accelerated SCI framework designed to overcome these bottlenecks. It first integrates a distributed, load-balanced global de-duplication algorithm to minimize redundancy and communication overhead at scale. To address compute limitations, it employs specialized, fine-grained CUDA kernels for exact coupled configuration generation. Finally, to break the single-GPU memory barrier exposed by this full acceleration, it incorporates a GPU memory-centric runtime featuring GPU-side pooling, streaming mini-batches, and overlapped offloading. This design enables much larger configuration spaces and shifts the bottleneck from host-side limitations back to on-device inference. Our evaluation demonstrates that our work fundamentally expands the scale of solvable problems. On an NVIDIA A100 cluster with 64 GPUs, our work achieves up to 2.32X end-to-end speedup over the highly-optimized NNQS-SCI baseline while preserving the same chemical accuracy. Furthermore, it demonstrates excellent distributed performance, maintaining over 90% parallel efficiency in strong scaling tests.

cs.DC↗

Clifford augmented density matrix renormalization group for \textit{ab initio} quantum chemistry

The recently proposed Clifford augmented density matrix renormalization group (CA-DMRG) method seamlessly integrates Clifford circuits with matrix product states, and takes advantage of the expression power from both. CA-DMRG has been shown to be able to achieve higher accuracy than standard DMRG on commonly used lattice models, with only moderate computational overhead compared to the latter. In this work, we propose an efficient scheme in CA-DMRG to deal with \textit{ab initio} quantum chemistry Hamiltonians, and apply it to study several molecular systems. Our numerical results show that CA-DMRG can reach higher accuracy than DMRG using the same bond dimension, pointing out a promising route to push the boundary of solving \textit{ab initio} quantum chemistry with strong static correlations.

quant-ph↗

NNQS-AFQMC: Neural network quantum states enhanced fermionic quantum Monte Carlo

We introduce an efficient approach to implement neural network quantum states (NNQS) as trial wavefunctions in auxiliary-field quantum Monte Carlo (AFQMC). NNQS are a recently developed class of variational ansätze capable of flexibly representing many-body wavefunctions, though they often incur a high computational cost during optimization. AFQMC, on the other hand, is a powerful stochastic projector approach for ground-state calculations, but it normally requires an approximate constraint via a trial wavefunction or trial density matrix, whose quality affects the accuracy. Recently it has been shown (Xiao et al, arXiv2505.18519) that a broad class of highly correlated wave-functions can be integrated into AFQMC through stochastic sampling techniques. In this work, we apply this approach and present a direct integration of NNQS with AFQMC, allowing NNQS to serve as high-quality trial wavefunctions for AFQMC with manageable computational cost. We test the NNQS-AFQMC method on the challenging nitrogen molecule (N$_2$) at stretched geometries. Our results demonstrate that AFQMC with an NNQS trial wavefunction can attain near-exact total energies, highlighting the potential of AFQMC with NNQS to overcome longstanding challenges in strongly correlated electronic structure calculations. We also outline future research directions for improving this promising methodology.

physics.chem-ph↗

Transformer-Based Neural Networks Backflow for Strongly Correlated Electronic Structure

Solving the electronic Schrödinger equation for strongly correlated systems remains one of the grand challenges in quantum chemistry. Here we demonstrate that Transformer architectures can be adapted to capture the complex grammar of electronic correlations through neural network backflow. In this approach, electronic configurations are processed as token sequences, where attention layers learn non-local orbital correlations and token-specific neural networks map these contextual representations into backflowed orbitals. Application to strongly correlated iron-sulfur clusters validates our approach: for $\left[\mathrm{Fe}_2 \mathrm{~S}_2\left(\mathrm{SCH}_3\right)_4\right]^{2-}$ ([2Fe-2S]) (30e,20o), the ground-state energy within chemical accuracy of DMRG while predicting magnetic exchange coupling constants closer to experimental values than all compared methods including DMRG, CCSD(T), and recent neural network approaches. For $\left[\mathrm{Fe}_4 \mathrm{S}_4\left(\mathrm{SCH}_3\right)_4\right]^{2-}$ ([4Fe-4S]) (54e,36o), we match DMRG energies and accurately reproduce detailed spin-spin correlation patterns between all Fe centers. The approach scales favorably to large active spaces inaccessible to exact methods, with distributed VMC optimization enabling stable convergence. These results establish Transformer-based backflow as a powerful variational ansatz for strongly correlated electronic structure, achieving superior magnetic property predictions while maintaining chemical accuracy in total energies.

quant-ph↗

SwarmThinkers: Learning Physically Consistent Atomic KMC Transitions at Scale

Can a scientific simulation system be physically consistent, interpretable by design, and scalable across regimes--all at once? Despite decades of progress, this trifecta remains elusive. Classical methods like Kinetic Monte Carlo ensure thermodynamic accuracy but scale poorly; learning-based methods offer efficiency but often sacrifice physical consistency and interpretability. We present SwarmThinkers, a reinforcement learning framework that recasts atomic-scale simulation as a physically grounded swarm intelligence system. Each diffusing particle is modeled as a local decision-making agent that selects transitions via a shared policy network trained under thermodynamic constraints. A reweighting mechanism fuses learned preferences with transition rates, preserving statistical fidelity while enabling interpretable, step-wise decision making. Training follows a centralized-training, decentralized-execution paradigm, allowing the policy to generalize across system sizes, concentrations, and temperatures without retraining. On a benchmark simulating radiation-induced Fe-Cu alloy precipitation, SwarmThinkers is the first system to achieve full-scale, physically consistent simulation on a single A100 GPU, previously attainable only via OpenKMC on a supercomputer. It delivers up to 4963x (3185x on average) faster computation with 485x lower memory usage. By treating particles as decision-makers, not passive samplers, SwarmThinkers marks a paradigm shift in scientific simulation--one that unifies physical consistency, interpretability, and scalability through agent-driven intelligence.

cs.AI↗

Analytical gradients of random-phase approximation plus corrections from renormalized single excitations

The random-phase approximation (RPA) formulated within the adiabatic connection fluctuation-dissipation framework is a powerful approach to compute the ground-state energies and properties of molecules and materials. Its overall underbinding behavior can be effectively mitigated by a simple correction term, called renormalized single excitation (rSE) correction. Analytical gradient calculations of the RPA energy have become increasingly available, enabling structural relaxations and even molecular dynamics at the RPA level. However, such calculations at the RPA+rSE level have not been reported, due to the lack of the rSE analytical gradient. Here, we present the first formulation and implementation of the analytical gradients of the rSE energy with respect to the nuclear coordinates within an atomic-orbital basis set framework, which allows us to assess the performance of RPA+rSE in determining the molecular geometries and energetics. It is found that the slight overestimation behavior of RPA for small covalently bonded molecules is strengthened by rSE, while such behavior for molecules bonded with purely dispersion interactions is corrected. We further applied the approach to the water clusters, and found that the energy difference between the low-energy isomers of water hexamers is almost unchanged when going from RPA to RPA+rSE geometries. For the bigger WATER27 test set, using the RPA+rSE geometries instead of the RPA ones leads to a slight reduction of the mean absolute error of RPA+rSE from 0.91 kcal/mol to 0.70 kcal/mol, at the complete basis set.

physics.chem-ph↗

Density-Functional Perturbation Theory with Numeric Atom-Centered Orbitals

This paper represents one contribution to a larger Roadmap article reviewing the current status of the FHI-aims code. In this contribution, the implementation of density-functional perturbation theory in a numerical atom-centered framework is summarized. Guidelines on usage and links to tutorials are provided.

cond-mat.mtrl-sci↗

Advancing Nonadiabatic Molecular Dynamics Simulations for Solids: Achieving Supreme Accuracy and Efficiency with Machine Learning

Non-adiabatic molecular dynamics (NAMD) simulations have become an indispensable tool for investigating excited-state dynamics in solids. In this work, we propose a general framework, N$^2$AMD which employs an E(3)-equivariant deep neural Hamiltonian to boost the accuracy and efficiency of NAMD simulations. The preservation of Euclidean symmetry of Hamiltonian enables N$^2$AMD to achieve state-of-the-art performance. Distinct from conventional machine learning methods that predict key quantities in NAMD, N$^2$AMD computes these quantities directly with a deep neural Hamiltonian, ensuring supreme accuracy, efficiency, and consistency. Furthermore, N$^2$AMD demonstrates excellent generalizability and enables seamless integration with advanced NAMD techniques and infrastructures. Taking several extensively investigated semiconductors as the prototypical system, we successfully simulate carrier recombination in both pristine and defective systems at large scales where conventional NAMD often significantly underestimates or even qualitatively incorrectly predicts lifetimes. This framework not only boosts the efficiency and precision of NAMD simulations but also opens new avenues to advance materials research.

physics.comp-ph↗

Rapidly Achieving Chemical Accuracy with Quantum Computing Enforced Language Model

Finding accurate ground state energy of a many-body system has been a major challenge in quantum chemistry. The integration of classic and quantum computers has shed new light on resolving this outstanding problem. Here we propose QiankunNet-VQE, a transformer based language models enforced with quantum computing to learn and generate quantum states. It has been implemented using up to 12 qubits and attaining an accuracy level competitive with state-of-the-art classical methods. By leveraging both quantum and classical resources, this scheme overcomes the limitations of variational quantum eigensolver(VQE) without the need for cumbersome error mitigation. Moreover, QiankunNet-VQE provides a different route to achieve a practical quantum advantage for solving many-electron Schrödinger equation without requiring extremely precise preparation and measurement of the ground-state wavefunction on quantum computer.

quant-ph↗

Efficient structural relaxation based on the random phase approximation: Applications to the water clusters

We report an improved implementation for evaluating the analytical gradients of the random phase approximation (RPA) electron-correlation energy based on atomic orbitals and the localized resolution of identity scheme. The more efficient RPA force calculations allow us to relax structures of medium-size water clusters. Particular attention is paid to the structures and energy orderings of the low-energy isomers of (H$_2$O)$_n$ clusters with $n=21$, 22, and 25. It is found that the energy ordering of the low-energy isomers of these water clusters are rather sensitive to how their structures are determined. For the five low-energy isomers of (H$_2$O)$_{25}$, the RPA energy ordering based on the RPA geometries is quite different from that based on the geometries relaxed by lower-level theories, in contrast with the situation of small water clusters like the water hexamer. The standard RPA underbinds the water clusters, and this underbinding behavior gets more pronounced as the complete basis set (CBS) limit is approached. The renormalized single excitation (rSE) correction remedies this underbinding, giving rise to a noticeable overbinding behavior at finite basis sets. However, as the CBS limit is approached, RPA+rSE yields an accuracy for the binding energies that is comparable to the best available double hybrid functionals, as demonstrated for the WATER27 testset.

cond-mat.mtrl-sci↗

Solving Schrödinger Equation with a Language Model

Accurately solving the Schrödinger equation for intricate systems remains a prominent challenge in physical sciences. A paradigm-shifting approach to address this challenge involves the application of artificial intelligence techniques. In this study, we introduce a machine-learning model named QiankunNet, based on the transformer architecture employed in language models. By incorporating the attention mechanism, QiankunNet adeptly captures intricate quantum correlations, which enhances its expressive power. The autoregressive attribute of QiankunNet allows for the adoption of an exceedingly efficient sampling technique to estimate the total energy, facilitating the model training process. Additionally, performance of QiankunNet can be further improved via a pre-training process. This work not only demonstrates the power of artificial intelligence in quantum mechanics but also signifies a pivotal advancement in extending the boundary of systems which can be studied with a full-configuration-interaction accuracy.

quant-ph↗