Fast point cloud generation with diffusion models in high energy physics

Many particle physics datasets like those generated at colliders are described by continuous coordinates (in contrast to grid points in an image), respect a number of symmetries (like permutation invariance), and have stochastic dimensionality. For this reason, standard deep generative models that produce images or least fixed set features limiting. We introduce new neural network simulation based on diffusion model addresses these limitations named fast point cloud diffusion. show our approach can reproduce the complex properties hadronic jets from proton-proton collisions with competitive precision other recently proposed models. Additionally, we use procedure called progressive distillation accelerate generation time method, which is typically significant challenge for despite their state-of-the-art precision.