A powerful procedure for optimizing AGP states

We present a powerful iterative algorithm for optimizing the geminal coefficients of an antisymmetrized geminal power (AGP) wavefunction. The algorithm is based on the lowest eigenvector of a matrix closely related to the Hessian of the problem. This matrix can be derived either by using Euler's theorem or by utilizing a unitary group approach. Two important features of the scheme are that it will always converge towards a minimum and that in the neighborhood of a minimum it is comparable to a quadratically convergent Newton‐Raphson method.