Five times faster CMB lensing measurements—here's how

Computing the realization-dependent bias in CMB lensing measurements is painfully slow—it requires running thousands of realistic simulations to get right. Namikawa and Sherwin sidestep this by comparing masked simulations against analytically solvable isotropic ones, differencing away the bias instead of calculating it directly. The approach cuts total computational cost by five-fold for near-future experiments and slots directly into existing analysis pipelines.