Neural networks boost sensitivity to rare Higgs decays by 40%

Physicists trained deep neural networks on simulated LHC collisions to spot a rare decay where a heavy Higgs turns into a muon and tau. The approach cuts the upper limits on this forbidden process by 36–46% compared to standard techniques, with visible mass of the decay products emerging as the key tell-tale sign.