Abstract:Neural network-based variational Monte Carlo (NN-VMC) has emerged as a promising cutting-edge technique of ab initio quantum chemistry. However, the high computational cost of existing approaches hinders their applications in realistic chemistry problems. Here, we report the development of a new NN-VMC method that achieves a remarkable speed-up by more than one order of magnitude, thereby greatly extending the applicability of NN-VMC to larger systems. Our key design is a novel computational framework named Forward Laplacian, which computes the Laplacian associated with neural networks, the bottleneck of NN-VMC, through an efficient forward propagation process. We then demonstrate that Forward Laplacian is not only versatile but also facilitates more developments of acceleration methods across various aspects, including optimization for sparse derivative matrix and efficient neural network design. Empirically, our approach enables NN-VMC to investigate a broader range of atoms, molecules and chemical reactions for the first time, providing valuable references to other ab initio methods. The results demonstrate a great potential in applying deep learning methods to solve general quantum mechanical problems.

What problem does this paper attempt to address?

The paper aims to address the high computational cost of applying Neural Network Variational Monte Carlo (NN-VMC) methods to practical chemical problems. Specifically, the goals of the paper can be summarized as follows: 1. **Improve Computational Efficiency**: Existing NN-VMC methods have very high computational costs when dealing with large-scale systems, which limits their application scope. The paper proposes a new computational framework—Forward Laplacian—that can significantly accelerate the computation process, especially in the application to large systems. 2. **Optimize Bottleneck Computation**: In NN-VMC methods, computing the Laplacian operator related to the neural network input is the most time-consuming part of the entire process. Traditional computation methods obtain the Laplacian operator indirectly by computing the Hessian matrix, which is inefficient. The Forward Laplacian method proposed in the paper directly computes the Laplacian operator, avoiding unnecessary computational steps. 3. **Expand Applicability**: By improving computational efficiency, the NN-VMC method can be applied to larger-scale studies of atoms, molecules, and chemical reactions, providing new possibilities for quantum chemistry research. 4. **Design Efficient Neural Network Structures**: The paper also proposes a new neural network structure called LapNet, which can better leverage the advantages of the Forward Laplacian method. Additionally, by introducing sparse derivative attention blocks, the computational efficiency is further improved. 5. **Relative Energy Estimation**: The paper not only focuses on the estimation of absolute energy but also pays special attention to the accurate calculation of relative energy, which is important for understanding chemical reaction mechanisms and catalyst design. The effectiveness of the proposed method in relative energy estimation is validated through experiments. In summary, the paper primarily addresses the issue of low computational efficiency in NN-VMC methods. By proposing a new computational framework and network structure design, it greatly expands the application scope of these methods, achieving significant progress, especially in relative energy calculations.

Forward Laplacian: A New Computational Framework for Neural Network-based Variational Monte Carlo

A computational framework for neural network-based variational Monte Carlo with Forward Laplacian

Interatomic Force from Neural Network Based Variational Quantum Monte Carlo

Penalty and auxiliary wave function methods for electronic Excitation in neural network variational Monte Carlo

Variational Monte Carlo on a Budget -- Fine-tuning pre-trained Neural Wavefunctions

Variance extrapolation method for neural-network variational Monte Carlo

Transferable Neural Wavefunctions for Solids

Wasserstein Quantum Monte Carlo: A Novel Approach for Solving the Quantum Many-Body Schrödinger Equation

Provable Convergence of Variational Monte Carlo Methods

Faster and lower scaling orbital-space Variational Monte Carlo

Highly Accurate Real-space Electron Densities with Neural Networks

Neural Projected Quantum Dynamics: a systematic study

Improved Optimization for the Neural-network Quantum States and Tests on the Chromium Dimer

Polynomially efficient quantum enabled variational Monte Carlo for training neural-network quantum states for physico-chemical applications

DeepQMC: an open-source software suite for variational optimization of deep-learning molecular wave functions

Neural network backflow for ab-initio quantum chemistry

NNQS-Transformer: an Efficient and Scalable Neural Network Quantum States Approach for Ab initio Quantum Chemistry

Enhancing the Expressivity of Variational Neural, and Hardware-Efficient Quantum States Through Orbital Rotations

Towards the ground state of molecules via diffusion Monte Carlo on neural networks

Ab initio quantum chemistry with neural-network wavefunctions

A Blocked Linear Method for Optimizing Large Parameter Sets in Variational Monte Carlo