We developed a Transformer-based method for positron track reconstruction in the MEG II experiment. The model acts as a hit classifier to remove pileup hits in the MEG II drift chamber, which operates under a high pileup occupancy of 35 % to 50 %. The trained model significantly improved hit purity, leading to enhancements in tracking efficiency and resolution by 15 % and 5 %, respectively, at a muon stopping rate of 5 × 1 0 7 μ / s e c . This improvement translates into an approximately 10 % increase in the sensitivity of the μ → e γ branching ratio measurement.
Dispoto et al. (Wed,) studied this question.