IRIS-HEP Fellow: Alan Chen



Fellowship dates: Jul – Aug, 2026

Home Institution: University of Michigan


Project: Logic Gate Neural Networks for Jet Substructure Classification

This project investigates logic neural networks (LNNs) for low-latency jet-substructure classification in high-energy physics. Using torchlogix, it will establish a reproducible workflow that trains differentiable logic networks, discretizes them into Boolean circuits, and validates compiled inference against the discrete software model. The study will evaluate trade-offs among classification accuracy, Boolean-circuit complexity, and inference latency to support hardware-efficient deployment in trigger-oriented workflows.



More information: My project proposal

Mentors:

  • Liv Våge (Princeton University)

  • Lino Gerlach (Princeton University)

Presentations and Publications
Current Status


Contact me: