Lattice Gauge Equivariant Convolutional Neural Networks

Favoni, Matteo; Ipp, Andreas; Müller, David I.; Schuh, Daniel

doi:10.1103/physrevlett.128.032003

Cited by 46 publications

(38 citation statements)

References 80 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We apply our method to a heuristic atmospheric weather model using using the NG-RC [13] as the core learning machine, which reduces the computational complexity and the data set size required for training while displaying state-of-the-art accuracy. Accounting for the translational symmetry of the weather model further reduces the training time and data, highlighting the importance of addressing symmetries [17][18][19][20][21]. Figure 1 illustrates our scheme.…”

mentioning

confidence: 99%

Learning Spatiotemporal Chaos Using Next-Generation Reservoir Computing

Barbosa¹,

Gauthier²

2022

Preprint

View full text Add to dashboard Cite

Forecasting the behavior of high-dimensional dynamical systems using machine learning (ML) requires efficient methods to learn the underlying physical model. We demonstrate spatiotemporal chaos prediction of a heuristic atmospheric weather model using an ML architecture that, when combined with a next-generation reservoir computer, displays state-of-the-art performance with a training time 10 3 − 10 4 times faster and training data set ∼ 10 2 times smaller than other ML algorithms. We also take advantage of the translational symmetry of the model to further reduce the computational cost and training data, each by a factor of ∼10.

show abstract

mentioning

confidence: 99%

Learning Spatiotemporal Chaos Using Next-Generation Reservoir Computing

Barbosa¹,

Gauthier²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…ML methods are being developed to speed up all three steps in such calculations [35]: (i) generation of gauge configurations (see Sec. 6), (ii) measurement of correlation functions [36,37,38,39] (see example below), and (iii) their analysis to extract physics [40,41,42]. The first two steps in LQCD, and ML systems, are computational and are therefore labeled "black boxes".…”

Section: Lattice Qcdmentioning

confidence: 99%

“…In ML too, one can impose symmetries at the level of the model (a penalty for models that break it or build it into the tuning of weights) or data (data representation itself builds in the symmetry or perform data augmentation to realize the symmetry). Ongoing work suggests that incorporating an understanding of the science or the dynamics of the system into ML is likely to dramatically increase its power [36,37,22,48].…”

Section: Contrasting Two Black Boxes: Lattice Qcd and MLmentioning

confidence: 99%

AI and Theoretical Particle Physics

Gupta¹,

Bhattacharya²,

Yoon³

2022

Preprint

View full text Add to dashboard Cite

Theoretical particle physicists continue to push the envelope in both high performance computing and in managing and analyzing large data sets. For example, the goals of sub-percent accuracy in predictions of quantum chromodynamics (QCD) using large scale simulations of lattice QCD and in finding signals of rare events and new physics in exabytes of data produced by experiments at the high luminosity large hadron collider (LHC) require new tools beyond just developments in hardware. Machine learning and artificial intelligence offer the promise of dramatically reducing the computational cost and time. This chapter reviews selected areas where AI/ML tools could have a major impact, provides an overview of the challenges, and discusses how new ideas such as normalizing flows can speed up the generation of gauge configurations needed in lattice QCD calculations; the growth of ML in surrogate models and pattern matching to reduce the cost of event generators and in the analysis of experimental data; and in the search for viable vacua in the landscape of string theories. While such approaches transform aspects of particle theory into computational problems, and thus black boxes, we argue that physics-aware development of these tools combined with algorithms that ensure that the results are bias free will continue to require a deep understanding of the physics. We see this broader transformation as akin to formulating and extracting observables from simulations of lattice QCD, a numerical integration of the path integral formulation of QCD that nevertheless requires a deep understanding of the underlying quantum field theory, the standard model of particle physics and effective field theory methods.

show abstract

“…Observable measurement -Quantities like correlation functions are evaluated over ensembles of field configurations. ML applications thus far include novel methods to extract thermodynamic observables [44], action parameter regression [67], observable approximation [68][69][70], design of new observables [58,[71][72][73][74][75][76][77][78][79][80], and path-integral contour deformations for baryonic correlators [81].…”

Section: Introductionmentioning

confidence: 99%

Applications of Machine Learning to Lattice Quantum Field Theory

Boyda¹,

Calì²,

Foreman³

et al. 2022

Preprint

View full text Add to dashboard Cite

There is great potential to apply machine learning in the area of numerical lattice quantum field theory, but full exploitation of that potential will require new strategies. In this white paper for the Snowmass community planning process, we discuss the unique requirements of machine learning for lattice quantum field theory research and outline what is needed to enable exploration and deployment of this approach in the future.

show abstract

Lattice Gauge Equivariant Convolutional Neural Networks

Cited by 46 publications

References 80 publications

Learning Spatiotemporal Chaos Using Next-Generation Reservoir Computing

Learning Spatiotemporal Chaos Using Next-Generation Reservoir Computing

AI and Theoretical Particle Physics

Applications of Machine Learning to Lattice Quantum Field Theory

Contact Info

Product

Resources

About