Speaker
Raghav Kansal
Description
Fast, accurate simulations and reconstruction are becoming increasingly necessary for precision measurements and searches planned in HL-LHC and beyond. Recent breakthroughs in deep generative modeling in computer vision and natural language processing offer a promising avenue for improving the speed and precision of future simulation paradigms, while providing a natural means for differentiable, end-to-end future detector optimization. We discuss state-of-the-art graph- and attention-based models for simulating and reconstructing LHC-like data, and the outlook for improving computing and detector design optimization for the FCC.
