This is SParTA: A modular, efficient and scalable transformer architecture for jet-tagging

Key points

• Exploiting the steeply falling transverse-momentum ($pT$) spectrum of jet constituents, SParTA factorizes constituent processing into two topological regimes.1
• Pairwise interaction features are computed dynamically in-network from constituent four-momenta and conditioned on global jet-level substructure observables.1
• On the quark-gluon tagging benchmark, a 300K-parameter SParTA model achieves an AUC of 92.02% and accuracy of 85.06%, matching the 2.14M-parameter Particle Transformer (ParT) at one-seventh the parameter count and outperforming IAFormer by 21% in background rejection at 30% signal efficiency.1
• On the boosted top-tagging benchmark, SParTA outperforms ParT and ParticleNet with 511K parameters, while we analyze the architectural trade-offs relative to fully dense multi-layer interaction networks.1

Sources (1)

• [1]This is SParTA: A modular, efficient and scalable transformer architecture for jet-tagging
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 19, 06:06 PM

  We present the Sparse Particle-interaction Transformer Architecture (SParTA), a modular, scalable, and efficient transformer for jet tagging at high-energy colliders.

  Exploiting the steeply falling transverse-momentum ($p_T$) spectrum of jet constituents, SParTA factorizes constituent processing into two topological regimes.

Extractive summary: sentences quoted from the sources.