Scalable variational Monte Carlo with graph neural ansatz
arXiv:2011.12453
Abstract
Deep neural networks have been shown as a potentially powerful ansatz in variational Monte Carlo for solving quantum many-body problems. We propose two improvements in this direction. The first is graph neural ansatz (GNA), which is a variational wavefunction universal to arbitrary geometry. GNA results in accurate ground-state energies on 2D Kagome lattices, triangular lattices, and randomly connected graphs. Secondly, we design a distributed workflow on multiple accelerators to scale up the computation. We compute Kagome lattices with sizes up to 432 sites on 128 TPU cores. The parameter sharing nature of the GNA also leads to transferability across different system sizes and geometries.
9 pages, 3 figures, NeurIPS 2020 workshop on Machine Learning and the Physical Sciences
References in corpus (21)
- Geometric deep learning: going beyond Euclidean data
- Solving the Quantum Many-Body Problem with Artificial Neural Networks
- Gapless Spin-Fluid Ground State in a Random Quantum Heisenberg Magnet
- Learning Combinatorial Optimization Algorithms over Graphs
- Layer Normalization
- Learning to Simulate Complex Physics with Graph Networks
- Supersolid hardcore bosons on the triangular lattice
- Restricted-Boltzmann-Machine Learning for Solving Strongly Correlated Quantum Systems
- Supersolid Order from Disorder: Hard-Core Bosons on the Triangular Lattice
- Graph networks as learnable physics engines for inference and control
- Deep autoregressive models for the efficient variational simulation of many-body quantum systems
- Persistent supersolid phase of hard-core bosons on the triangular lattice
- Study of the Two-Dimensional Frustrated J1-J2 Model with Neural Network Quantum States
- Approximating quantum many-body wave-functions using artificial neural networks
- Valence Bond Solids and Their Quantum Melting in Hard-Core Bosons on the Kagome Lattice
- Machine learning technique to find quantum many-body ground states of bosons on a lattice
- Solving frustrated quantum many-particle models with convolutional neural networks
- Chiral Topological Phases from Artificial Neural Networks
- Continuous easy-plane deconfined phase transition on the kagome lattice
- Decoding Molecular Graph Embeddings with Reinforcement Learning
- Variational optimization in the AI era: Computational Graph States and Supervised Wave-function Optimization
Cited by in corpus (4)
- Determinant-free fermionic wave function using feed-forward neural networks
- Group Convolutional Neural Networks Improve Quantum State Accuracy
- Rayleigh-Gauss-Newton optimization with enhanced sampling for variational Monte Carlo
- Ab-Initio Potential Energy Surfaces by Pairing GNNs with Neural Wave Functions