EINN2025

First Determination of the Collins–Soper Kernel using Lattice QCD in a Neural Network TMD fit

Not scheduled

20m

Talk in workshop 2: "AI & ML in nuclear science: starting with design, optimization, and operation of the machine and detectors, to data analysis"

Speaker

Chiara Bissolotti (Argonne National Laboratory)

Description

We present the first proof of concept extraction using neural networks (NNs) of the unpolarized transverse-momentum distributions (TMDs) at next-to-next-to-next-to-leading logarithmic (N$^3$LL) accuracy. By offering a more flexible and adaptable approach, NNs overcome some of the limitations of traditional functional forms, providing a better description of data.
Moreover, we present the first joint study of the Collins–Soper kernel combining inputs from lattice QCD and TMD phenomenology. Using recent continuum-extrapolated lattice calculations of the kernel at 3 values of the lattice spacing, we assess their impact on a recent phenomenological extraction based on Neural Network parametrizations. We perform both Bayesian reweighing and, for the first time, a direct global fit including the 21 lattice data alongside about 500 experimental measurements. We
find that the inclusion of the lattice points shifts the central value of the non-perturbative parameter by 5\% and reduces its uncertainty by 30\%, highlighting the potential of lattice inputs to improve TMD extractions.