Fragmentation Functions for $\Xi ^-/\bar{\Xi}^+$ Using Neural Networks

TL;DR

This paper introduces neural network-based fragmentation functions for the $\\Xi^-/\bar{\Xi}^+$ baryon, determined at NNLO, improving theoretical predictions for baryon production in high-energy collisions.

Contribution

It provides the first NNLO fragmentation functions for $\\Xi^-/\bar{\Xi}^+$ baryons using neural networks, enhancing the accuracy of QCD predictions.

Findings

Fragmentation functions determined at NNLO accuracy.

Improved fit quality with higher-order QCD corrections.

Predictions for baryon production at LHC energies.

Abstract

We present a determination of fragmentation functions (FFs) for the octet baryon $\Xi ^-/\bar{\Xi}^+$ from data for single inclusive electron-positron annihilation. Our parametrization in this QCD analysis is provided in terms of a Neural Network (NN). We determine fragmentation functions for $\Xi ^-/\bar{\Xi}^+$ at next-to-leading order and for the first time at next-to-next-to-leading order in perturbative QCD. We discuss the improvement of higher-order QCD corrections, the quality of fit, and the comparison of our theoretical results with the fitted datasets. As an application of our new set of fragmentation functions, named SHKS22, we present predictions for $\Xi ^- / \bar{\Xi}^+$ baryon production in proton-proton collisions at the LHC experiments.