 Informant function

The score is the gradient of the log-likelihood, $S_n(\theta )={\nabla }_{\theta }{\ell }_n(\theta )$.