Rare Event Analysis via Stochastic Optimal Control
Researchers have developed a new framework that frames the estimation of committor functions as a stochastic optimal control problem. This approach uses a feedback control mechanism, derived from the gradient of the committor's logarithm, to efficiently guide simulations toward rare events. The method includes strategies to overcome metastability and improve sampling, demonstrating superior accuracy in estimating reaction rates and equilibrium constants on benchmark systems compared to existing techniques. AI