A simplified eigenvector-following technique for locating transition points in an energy landscape

We derive an eigenvector-following technique for locating transition points in an N-dimensional energy landscape. A separate Lagrange multiplier is used for each eigendirection to provide maximum flexibility in determining step sizes. In contrast to previous techniques based on a similar approach, we provide a simple algorithm for choosing specific values of these Lagrange multipliers. We demonstrate the robustness of the algorithm using two-dimensional Cerjan-Miller and Adams landscapes. The technique has also been applied to the S-12 molecular cluster.