Top Quark

Top quark identification is crucial in high-energy physics experiments, aiming to accurately classify and measure properties of these massive particles within complex collision events. Current research focuses on developing efficient and interpretable machine learning models, such as permutation-equivariant and Lorentz-invariant neural networks (e.g., PELICAN) and graph neural networks (e.g., incorporating HaarPooling), to improve the accuracy and speed of top quark tagging and momentum reconstruction. These advancements leverage underlying physics symmetries to reduce model complexity and enhance performance, significantly impacting data analysis in particle physics experiments. The resulting improvements in efficiency and accuracy are vital for analyzing the massive datasets generated by modern particle accelerators.