Obtaining a physically well-founded theoretical description of hadrons across energy scales is of high importance for an intrinsic understanding of QCD and the extraction of precision quantities such as hadronic contributions to the Lamb shift or the impact of neutrino-nucleon interactions on neutrino oscillation experiments.When probing hadrons with electron beams, at high energies the underlying physics is well understood in terms of perturbative QCD. At low energies, the connection to the physics of the quark and gluon constituents becomes obscured and it is justified to use effective field theories that directly describe hadron degrees of freedom.In addition, to analytically connect the low and high-energy regimes, the wealth of resonances appearing in the spectrum needs to be accounted for, whose description is highly convoluted. Many of them, the exotic resonances, do not even follow the usual 2 or 3-valence-quark picture.I present an overview of the advances on obtaining a theoretically well founded understanding of the proton structure across energy scales, digressing also about the power and (current) limitations of machine learning as a tool towards this end.
Paulo Brás, Paulo Silva, Jaime Silva