Abstract
Following Yulong's talk, I will continue the discussion on birth-death sampling dynamics. I will first mention the global long-time convergence result of the idealized dynamics, which improves on the result of previous work and makes the dynamics appealing for sampling distributions with nonconvex potentials. After that, I will discuss what we currently understand and the remaining questions about the numerical method for its implementation - the blob method introduced by Carrillo, Craig and Patacchini. This is joint ongoing work with Yulong Lu and Dejan Slepcev.