Abstract: Employing the Bethe-Salpeter formalism, we present a computation of the space-like baryonic form factor for the pion and kaon. In the exact isospin-symmetric limit this observable is forbidden by G-parity, so that any nonzero signal constitutes a direct probe of the quark mass difference m(d) – m(u). The form factors are evaluated in the impulse approximation using fully dressed quark propagators, meson Bethe-Salpeter amplitudes, and a dressed baryon-current vertex constrained by the vector Ward-Takahashi identity. The baryonic radius computed with this method for the pion is given by < r(B)(2)>(1/2)(pi+) = 0.043(2) fm, and is consistent with the available dispersive benchmarks. Our predictions for the kaons, namely < r(B)(2)>(1/2)(K+) = 0.265(7)fm and < r(B)(2)>(1/2)(K0) = 0.262(7) fm, indicate a larger spatial extent than in the pion case; these results have no dispersive counterparts, and are compatible with chiral QCD models.

Abstract: Measurements of the total and differential Higgs boson production cross-sections, via WH and ZH associated production using H -> WW* -> lvlv and H -> WW*. lvjj decays, are presented. The analysis uses proton-proton events delivered by the Large Hadron Collider at a centre-of-mass energy of 13TeV and recorded by the ATLAS detector between 2015 and 2018. The data correspond to an integrated luminosity of 140 fb(-1). The sum of the WH and ZH cross-sections times the H -> WW* branching fraction is measured to be 0.44(-0.09)(+0.10) (stat.) (+0.06)(-0.05) (syst.) pb, in agreement with the Standard Model prediction. Higgs boson production is further characterised through measurements of the differential cross-section as a function of the transverse momentum of the vector boson and in the framework of Simplified Template Cross-Sections.

Abstract: Symmetries play a central role in physics, organizing dynamics, constraining interactions, and determining the effective number of physical degrees of freedom. In parallel, modern artificial intelligence methods have demonstrated a remarkable ability to extract low-dimensional structure from high-dimensional data through representation learning. This review examines the interplay between these two perspectives, focusing on the extent to which symmetry-induced constraints can be identified, encoded, or diagnosed using machine learning techniques. Rather than emphasizing architectures that enforce known symmetries by construction, we concentrate on data-driven approaches and latent representation learning, with particular attention to variational autoencoders. We discuss how symmetries and conservation laws reduce the intrinsic dimensionality of physical datasets, and how this reduction may manifest itself through self-organization of latent spaces in generative models trained to balance reconstruction and compression. We review recent results, including case studies from simple geometric systems and particle physics processes, and analyze the theoretical and practical limitations of inferring symmetry structure without explicit inductive bias.