Efficient Anti-Symmetrization of a Neural Network Layer by Taming the Sign Problem
Journal of Machine Learning · 2023
DOI: 10.4208/jml.230703 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Machine Learning, 2023. Verfasst von Nilin Abrahamsen und 1 weiteren Autor:innen. Die Arbeit nennt 17 Literaturangaben.
Abstract (Originalsprache)
Explicit antisymmetrization of a neural network is a potential candidate for a universal function approximator for generic antisymmetric functions, which are ubiquitous in quantum physics. However, this procedure is a priori factorially costly to implement, making it impractical for large numbers of particles. The strategy also suffers from a sign problem. Namely, due to near-exact cancellation of positive and negative contributions, the magnitude of the antisymmetrized function may be significantly smaller than before antisymmetrization. We show that the anti-symmetric projection of a two-layer neural network can be evaluated efficiently, opening the door to using a generic anti-symmetric layer as a building block in anti-symmetric neural network Ansatzes. This approximation is effective when the sign problem is controlled, and we show that this property depends crucially the choice of activation function under standard Xavier/He initialization methods. As a consequence, using a smooth activation function requires rescaling of the neural network weights compared to standard initializations.
Zitieren
APA
Abrahamsen, N., & Lin, L. (2023). Efficient Anti-Symmetrization of a Neural Network Layer by Taming the Sign Problem. Journal of Machine Learning. https://doi.org/10.4208/jml.230703
Vancouver
Abrahamsen N, Lin L. Efficient Anti-Symmetrization of a Neural Network Layer by Taming the Sign Problem. Journal of Machine Learning. 2023. doi:10.4208/jml.230703
BibTeX
@article{Abrahamsen2023,
author = {Abrahamsen, Nilin and Lin, Lin},
title = {Efficient Anti-Symmetrization of a Neural Network Layer by Taming the Sign Problem},
journal = {Journal of Machine Learning},
year = {2023},
doi = {10.4208/jml.230703}
}
EUVIMED