Symmetry-aware neural networks for particle collision tracking
Seth Nabat · Winnetka, California
Tenth Place ($40,000) · Regeneron Science Talent Search · 2026
Neural networks for reconstructing particle collision tracks, built to respect the symmetries of the underlying physics rather than learning them from data.
Building known physics into the architecture instead of hoping the network infers it is the cheapest possible source of accuracy — it is prior knowledge that costs no extra data.
- Symmetry-equivariant network design
- Particle track reconstruction
- Collision data simulation
- Benchmarking against a physics baseline
- Encode what you already know into the model; do not make it rediscover physics from examples
- Physics-informed machine learning is judged against a physics baseline, not only against other networks
Archive entries are summaries compiled for study. Read them for how a project was structured, argued, and defended — not as a template to reproduce. Copying someone else's project is the one thing that will end yours.
Compiled by ResearchForge from the Society for Science public award announcement linked on this record. Names, hometowns, project titles and reported results are as published. The "why it worked" and "what to take from this" notes are ResearchForge's editorial reading of that public record — they are not statements by the students and not the judges' rationale. All rights to the original projects remain with their authors; no project materials are rehosted here.