Jeffreys Prior

It’s a bayesian correction, the $0.5$ that you see in the RSJ paper