The adoption of machine learning (ML) into theoretical physics comes on the heels of an explosion of industry progress that started in 2012. Since that time, computer scientists have demonstrated that learning algorithms - those designed to respond and adapt to new data - provide an exceptionally powerful platform for tackling many difficult tasks in image recognition, natural language comprehension, game play and more. This new breed of ML algorithm has now conquered benchmarks previously thought to be decades away due to their high mathematical complexity. In the last several years, researchers at Perimeter have begun to examine machine learning algorithms for application to a new set of problems, including condensed matter, quantum information, numerical relativity, quantum gravity and astrophysics.
The sign problem is a widespread numerical hurdle preventing us from simulating the equilibrium behaviour of many interesting models, most notably the Hubbard model. Research aimed at solving the sign problem, via various clever manipulations, has been thriving for a long time with various recent exciting results. The complementary question, of whether some phases of matter forbid the existence of any sign-free microscopic model, has received attention only recently.
Neural networks (NNs) normally do not allow any insight into the reasoning behind their predictions. We demonstrate how inﬂuence functions can unravel the black box of NN when trained to predict the phases of the one-dimensional extended spinless Fermi-Hubbard model at half-ﬁlling. Results provide strong evidence that the NN correctly learns an order parameter describing the quantum transition.
We propose a reinforcement learning (RL) scheme for feedback quantum control within the quantum approximate optimization algorithm (QAOA). QAOA requires a variational minimization for states constructed by applying a sequence of unitary operators, depending on parameters living in a highly dimensional space. We reformulate such a minimum search as a learning task, where a RL agent chooses the control parameters for the unitaries, given partial information on the system. We show that our RL scheme learns a policy converging to the optimal adiabatic solution for QAOA found by Mbeng et al.
So far artificial neural networks have been applied to discover phase diagrams in many different physical models. However, none of these studies have revealed any fundamentally new physics. A major problem is that these neural networks are mainly considered as black box algorithms. On the journey to detect new physics it is important to interpret what artificial neural networks learn. On the one hand this allows us to judge whether to trust the results, and on the other hand this can give us insight to possible new physics. In this talk I will
We implement projective quantum Monte Carlo (PQMC) methods to simulate quantum annealing on classical computers. We show that in the regime where the systematic errors are well controlled, PQMC algorithms are capable of simulating the imaginary-time dynamics of the Schroedinger equation both on continuous space models and discrete basis systems. We also demonstrate that the tunneling time of the PQMC method is quadratically faster than the one of incoherent quantum annealing.
We propose to generalise classical maximum likelihood learning to density matrices. As the objective function, we propose a quantum likelihood that is related to the cross entropy between density matrices. We apply this learning criterion to the quantum Boltzmann machine (QBM), previously proposed by Amin et al. We demonstrate for the first time learning a quantum Hamiltonian from quantum statistics using this approach. For the anti-ferromagnetic Heisenberg and XYZ model we recover the true ground state wave function and Hamiltonian.
Short-depth algorithms are crucial for reducing computational error on near-term quantum computers, for which decoherence and gate infidelity remain important issues. Here we present a machine-learning inspired approach for discovering such algorithms. We apply our method to a ubiquitous primitive: computing the overlap Tr(rho*sigma) between two quantum states rho and sigma. The standard algorithm for this task, known as the Swap Test, is used in many applications such as quantum support vector machines, and, when specialized to rho=sigma, quantifies the Renyi entanglement.