Abstract:Quantum many-body systems are of great interest for many research areas, including physics, biology and chemistry. However, their simulation is extremely challenging, due to the exponential growth of the Hilbert space with the system size, making it exceedingly difficult to parameterize the wave functions of large systems by using exact methods. Neural networks and machine learning in general are a way to face this challenge. For instance, methods like Tensor networks and Neural Quantum States are being investigated as promising tools to obtain the wave function of a quantum mechanical system. In this tutorial, we focus on a particularly promising class of deep learning algorithms. We explain how to construct a Physics-Informed Neural Network (PINN) able to solve the Schrödinger equation for a given potential, by finding its eigenvalues and eigenfunctions. This technique is unsupervised, and utilizes a novel computational method in a manner that is barely explored. PINNs are a deep learning method that exploits Automatic Differentiation to solve Integro-Differential Equations in a mesh-free way. We show how to find both the ground and the excited states. The method discovers the states progressively by starting from the ground state. We explain how to introduce inductive biases in the loss to exploit further knowledge of the physical system. Such additional constraints allow for a faster and more accurate convergence. This technique can then be enhanced by a smart choice of collocation points in order to take advantage of the mesh-free nature of the PINN. The methods are made explicit by applying them to the infinite potential well and the particle in a ring, a challenging problem to be learned by an Artificial Intelligence agent due to the presence of complex-valued eigenfunctions and degenerate states.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the challenge of simulating quantum many - body systems. Specifically, since the Hilbert space grows exponentially with the system size, it becomes extremely difficult to parameterize the wave functions of large systems using exact methods. Neural networks and machine learning offer a way to meet this challenge. This paper focuses in particular on a class of deep - learning algorithms - Physics - Informed Neural Networks (PINNs), and explains how to construct a PINN that can find its eigenvalues and eigenfunctions by solving the Schrödinger equation for a given potential. This method is unsupervised and utilizes a novel computational method that can solve integro - differential equations without relying on grids. In addition, the paper also explores how to introduce inductive biases in the loss function to take advantage of further knowledge of the physical system, thereby achieving faster and more accurate convergence. This technique is further enhanced by the intelligent selection of collocation points to fully utilize the mesh - free characteristics of PINNs. The paper illustrates in detail the application of these methods through two examples: the infinite potential well and the particle in a ring.

A Tutorial on the Use of Physics-Informed Neural Networks to Compute the Spectrum of Quantum Systems

Addressing the Non-perturbative Regime of the Quantum Anharmonic Oscillator by Physics-Informed Neural Networks

Physics informed neural networks learning a two-qubit Hamiltonian

On physics-informed neural networks for quantum computers

Neural networks in quantum many-body physics: a hands-on tutorial

How To Use Neural Networks To Investigate Quantum Many-Body Physics

Physics-Informed Neural Networks for Quantum Control

Physics-informed neural network for quantum control of NMR registers

Unsupervised Random Quantum Networks for PDEs

Physics-informed neural networks viewpoint for solving the Dyson-Schwinger equations of quantum electrodynamics

Physics-Informed Neural Networks as Solvers for the Time-Dependent Schrödinger Equation

Physics-Informed Neural Networks for an optimal counterdiabatic quantum computation

Solving the Regge-Wheeler and Teukolsky equations: supervised versus unsupervised physics-informed neural networks

Computation of the Time-Dependent Dirac Equation with Physics-Informed Neural Networks

Scientific Machine Learning Through Physics–Informed Neural Networks: Where we are and What's Next

Using physics-informed neural networks to compute quasinormal modes

Exploring Physics-Informed Neural Networks: From Fundamentals to Applications in Complex Systems

From {\tt Ferminet} to PINN. Connections between neural network-based algorithms for high-dimensional Schrödinger Hamiltonian