Jet flavor tagging with Particle Transformer for Higgs factories

We study the performance of the Particle Transformer (ParT) for jet flavor tagging using ILD full simulation events (1M jets) as well as fast simulation samples (10M and 1M jets). We perform 3-category ($b\mathrm{/}c\mathrm{/}d$), 6-category ($b\mathrm{/}c\mathrm{/}d\mathrm{/}u\mathrm{/}s\mathrm{/}g$), and 11-category trainings (including quark–antiquark separation), incorporating multivariate hadron particle identification information from $dE\mathrm{/}dx$ and time-of-flight. For $b$/$c$ tagging, we observe a factor of 5–10 improvement over previous BDT-based taggers, and we obtain reasonable performance for strange tagging and quark/antiquark separation. The 10M-jet fast simulation study indicates that further gains are possible with higher training statistics.