Speaker
Bingxuan Liu
(SunYat-sen University)
Description
The ATLAS collaboration has developed a new transformer-based flavour tagging algorithm, GN2. It achieves much better performance compared to the previous taggers. In addition, the development-deployment-calibration cycle has been greatly streamlined so that it can be quickly adapted to new use cases. GN2 has already been used in physics analyses, demonstrating its superior performance. Our Chinese institutes have played critical roles in this process. The paper has been accepted by Nature Communications, which should be published in time for CLHCP.
Primary author
Bingxuan Liu
(SunYat-sen University)
