arXiv:2310.08053 · Physical review. D/Physical review. D.

Lattice real-time simulations with learned optimal kernels

Daniel Alvestad, Alexander Rothkopf, Dénes Sexty

Published Feb 5, 2024• 11 citations

Abstract

We present a simulation strategy for the real-time dynamics of quantum fields, inspired by reinforcement learning. It builds on the complex Langevin approach, which it amends with system-specific prior information, a necessary prerequisite to overcome this exceptionally severe sign problem. The optimization process underlying our machine-learning approach is made possible by deploying inherently stable solvers of the complex Langevin stochastic process and a novel optimality criterion derived from insight into so-called boundary terms. This conceptual and technical progress allows us to both significantly extend the range of real-time simulations in 1 + 1 d scalar field theory beyond the state of the art and to avoid discretization artifacts that plagued previous real-time field theory simulations. Limitations of and promising future directions are discussed. Published by the American Physical Society 2024

DOI: 10.1103/physrevd.109.l031502

Related papers

Embedding memory-efficient stochastic simulators as quantum trajectories

Thomas J. Elliott, Mile Gu

By exploiting the complexity intrinsic to quantum dynamics, quantum technologies promise a host of computational advantages. One such advantage lies in the field of stochastic modeling, where it has been shown that quantum stochastic simulators can operate with a lower memory overhead than their best classical counterparts. This advantage is particularly pronounced for continuous-time stochastic processes; however, the corresponding quantum stochastic simulators heretofore prescribed operate only on a quasi-continuous-time basis and suffer an ever-increasing circuit complexity with increasing temporal resolution. Here, by establishing a correspondence with quantum trajectories (a method for modeling open quantum systems), we show how truly continuous-time quantum stochastic simulators can be embedded in such open quantum systems, bridging this gap and obviating previous constraints. We further show how such an embedding can be made for discrete-time stochastic processes, which manifest as jump-only trajectories, and discuss how viewing the correspondence in the reverse direction provides a means of studying structural complexity in quantum systems themselves. Published by the American Physical Society 2024

arXiv:2402.04708Physical Review AFeb 26, 20244 citations

Title: Non-Linear Quantum Field Theory with Adaptive Parameterization

Jincheng Zhang

This paper introduces a novel non-linear quantum field theory framework designed to enhance model accuracy and computational efficiency. Traditional field theories frequently rely on static, predefined parameters, limiting their applicability to specific physical scenarios. We propose a dynamic parameterization system that iteratively adjusts these parameters based on model performance, creating a feedback loop that continuously refines the model's fit. The core mechanism centers around a reinforcement learning-inspired feedback loop, evaluating the model's output against new data and optimizing parameter values to maximize predictive power. This approach overcomes the limitations of fixed parameter models, offering a more flexible and efficient framework for simulating complex quantum phenomena. The resulting model demonstrates improved performance across a range of benchmark problems, highlighting the potential for this adaptive approach in diverse physical contexts.

Sep 1, 20260 citations

Adaptive Quantum Computing Simulations for Machine Learning

Jincheng Zhang

The simulation of quantum computers presents a formidable challenge due to the exponential growth in computational cost with the number of qubits. Current simulation methods often struggle to scale to even moderately sized quantum systems, severely limiting their applicability for machine learning applications. This paper introduces an adaptive simulation strategy designed to mitigate this issue. The core idea revolves around dynamically adjusting the simulation complexity—the level of detail—based on the demands of the underlying machine learning task. We propose a hierarchical simulation approach, where a coarse-grained representation is utilized initially, and the simulation is progressively refined, or "zoomed in," only when specific regions or interactions become critical for the learning process. This strategy minimizes unnecessary computations, leading to a significant reduction in overall simulation time and resource requirements. The effectiveness of this approach is demonstrated through theoretical analysis and the formulation of key equations governing the adaptive simulation process. The primary goal is to achieve a more efficient and practical approach to quantum computing simulations for machine learning, paving the way for the exploration of quantum algorithms within the machine learning domain.

Sep 3, 20260 citations

Quantum-Inspired Algorithms for Reinforcement Learning

Jincheng Zhang

This paper investigates the application of quantum-inspired algorithms to the field of reinforcement learning (RL). Traditional RL methods often struggle with exploration challenges and can be computationally expensive, particularly in complex environments. We propose a novel RL algorithm that leverages concepts from quantum mechanics, specifically superposition and entanglement, to address these limitations. The core idea is to represent states, actions, and rewards using quantum-inspired representations, allowing the agent to explore the environment more efficiently and potentially converge to optimal policies faster. This work presents a detailed formulation of the algorithm, including the mathematical framework and key considerations for implementation. The potential benefits of this approach, including improved exploration and reduced computational complexity, are discussed. The algorithm is designed to iteratively refine its policy based on simulated quantum measurements, aiming to minimize the expected return. The exploration strategy is guided by the inherent probabilistic nature of quantum states, leading to a more diverse and potentially faster learning process. Further research is warranted to fully explore the capabilities and limitations of this approach across various RL benchmark problems.

Sep 1, 20260 citations

Core Claim: Develop a machine learning algorithm based on adaptive quantum topology field simulation, accelerating quantum machine learning algorithms and enhancing algorithm generalization.

Jincheng Zhang

Quantum machine learning is a rapidly evolving field with the potential to revolutionize various computational tasks. This paper investigates the application of adaptive quantum topology field simulation as a novel approach to enhance quantum machine learning. We propose a machine learning algorithm that dynamically adjusts the number and state of quantum bits, leveraging the unique properties of quantum topology to accelerate training and improve generalization. The core mechanism centers around a system of optimized quantum circuit design coupled with adaptive quantum state management. The development of this algorithm aims to overcome limitations of traditional quantum machine learning algorithms, particularly in terms of computational complexity and adaptability. This paper presents a detailed theoretical analysis and preliminary experimental results demonstrating the effectiveness of this approach in a range of quantum machine learning tasks. The research explores the potential benefits of this method in terms of faster training times, improved model accuracy, and enhanced robustness across diverse datasets. Finally, we discuss the challenges and future directions for further investigation in this field.

Sep 3, 20260 citations

Ground States of Quantum Many Body Lattice Models via Reinforcement Learning

Willem Gispen, Austen Lamacraft

We introduce reinforcement learning (RL) formulations of the problem of finding the ground state of a many-body quantum mechanical model defined on a lattice. We show that stoquastic Hamiltonians - those without a sign problem - have a natural decomposition into stochastic dynamics and a potential representing a reward function. The mapping to RL is developed for both continuous and discrete time, based on a generalized Feynman-Kac formula in the former case and a stochastic representation of the Schrödinger equation in the latter. We discuss the application of this mapping to the neural representation of quantum states, spelling out the advantages over approaches based on direct representation of the wavefunction of the system.

arXiv:2012.07063Dec 13, 20201 citation