Accurate computation of quantum excited states with neural networks

2023/08/31 by David Pfau, Simon Axelrod, Halvard Sutterud +2 · 1 voice · 50 citations
Materials Science · Mathematics · Physics and Astronomy · #Advanced Chemical Physics Studies #Algorithm #Artificial intelligence #Artificial neural network #Computation #Computer science #Diagonal #Dipole #Excitation #Excited state #Machine Learning in Materials Science #Mathematics #Observable #Orthogonalization #Physics #Quantum #Quantum mechanics #Spectroscopy and Quantum Chemical Studies #Statistical physics #Work (physics)

paper · pdf · doi:10.1126/science.adn0137

published in Science 385(6711), eadn0137 (American Association for the Advancement of Science)

openalex publication_date 2024/08/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/30

Abstract

We present an algorithm to estimate the excited states of a quantum system by variational Monte Carlo, which has no free parameters and requires no orthogonalization of the states, instead transforming the problem into that of finding the ground state of an expanded system. Arbitrary observables can be calculated, including off-diagonal expectations, such as the transition dipole moment. The method works particularly well with neural network ansätze, and by combining this method with the FermiNet and Psiformer ansätze, we can accurately recover excitation energies and oscillator strengths on a range of molecules. We achieve accurate vertical excitation energies on benzene-scale molecules, including challenging double excitations. Beyond the examples presented in this work, we expect that this technique will be of interest for atomic, nuclear, and condensed matter physics.

Citations

• A Stochastic Approximation Method
• Visible Light Photoredox Catalysis with Transition Metal Complexes: Applications in Organic Synthesis

Cited by

• Analyzing Atomic Interactions in Molecules as Learned by Neural Networks
• Nonplanar qubit with tunable gauge symmetry
• Ab-initio simulation of excited-state potential energy surfaces with transferable deep quantum Monte Carlo
• Accurate Ab-initio Neural-network Solutions to Large-Scale Electronic Structure Problems
• Dissipative quantum algorithms for excited-state quantum chemistry
• Revisiting the Broken Symmetry Phase of Solid Hydrogen: A Neural Network Variational Monte Carlo Study
• Fermionic neural Gibbs states
• A Global Spacetime Optimization Approach to the Real-Space Time-Dependent Schrödinger Equation
• Quantum-centric machine learning for molecular dynamics
• QMeCha: quantum Monte Carlo package for fermions in embedding environments
• Modal Backflow Neural Quantum States for Anharmonic Vibrational Calculations
• Spin-Adapted Neural Network Wavefunctions in Real Space
• Benchmarking Simulacra AI's Quantum Accurate Synthetic Data Generation for Chemical Sciences
• Convolutional transformer wave functions
• Revisiting Orbital Minimization Method for Neural Operator Decomposition
• Cross-disciplinary perspectives on the potential for artificial intelligence across chemistry
• A Variance-Based Convergence Criterion in Neural Variational Monte Carlo for Quantum Systems
• Adiabatic transport of neural network quantum states
• Bound on entanglement in neural quantum states
• Accurate ground states of SU(2) lattice gauge theory in 2+1D and 3+1D
• Simultaneous approximation of multiple degenerate states using a single neural network quantum state
• Towards Interpretability of Neural Quantum States
• Sum-of-Gaussians tensor neural networks for high-dimensional Schrödinger equation
• Machine learning the single-Λ hypernuclei with neural-network quantum states
• Neural Scaling Laws Surpass Chemical Accuracy for the Many-Electron Schrödinger Equation
• Neural Importance Resampling: A Practical Sampling Strategy for Neural Quantum States
• Solving excited states for long-range interacting trapped ions with neural networks
• Neural Network-Augmented Pfaffian Wave-functions for Scalable Simulations of Interacting Fermions
• Functional Neural Wavefunction Optimization
• Variational subspace methods and application to improving variational Monte Carlo dynamics
• A Neural-Guided Variational Quantum Algorithm for Efficient Sign Structure Learning in Hybrid Architectures
• Thermalization Dynamics in the Two-Dimensional Hubbard Model with Neural-Network Quantum States
• The QUEST Database of Highly-Accurate Excitation Energies
• QMCTorch: Molecular Wavefunctions with Neural Components for Energy and Force Calculations
• Imaginary time spectral transforms for excited-state preparation
• A Unified Variational Framework for Quantum Excited States

Discussions

• Do people post preprint threads on Bluesky? I guess they do now! Thrilled to share what I’ve been working on for the last two years - a new way to solve one of the most fundamental problems in quantum [bsky, 35 points, 1 comments]

Related

• Calculate electronic excited states using neural networks with effective core potential
• Approximating Excited States using Neural Networks
• Ab-initio simulation of excited-state potential energy surfaces with transferable deep quantum Monte Carlo
• Neural Quantum States in Variational Monte Carlo Method: A Brief Summary
• A Promising Intersection of Excited-State-Specific Methods from Quantum Chemistry and Quantum Monte Carlo