19 Parameters Is All You Need: Tiny Neural Networks for Particle Physics

TL;DR

This paper introduces ultra-lightweight neural networks with as few as 19 parameters, demonstrating their effectiveness in particle physics tasks like top quark jet tagging, suitable for low-latency applications.

Contribution

It presents minimal neural network architectures based on PELICAN that outperform larger models in particle physics classification tasks.

Findings

19-parameter models outperform larger architectures in jet tagging

Lightweight models enable faster, low-latency particle physics analysis

PELICAN architecture is effective with minimal parameters

Abstract

As particle accelerators increase their collision rates, and deep learning solutions prove their viability, there is a growing need for lightweight and fast neural network architectures for low-latency tasks such as triggering. We examine the potential of one recent Lorentz- and permutation-symmetric architecture, PELICAN, and present its instances with as few as 19 trainable parameters that outperform generic architectures with tens of thousands of parameters when compared on the binary classification task of top quark jet tagging.