Dear developers,
I recently used CI-NEB and read the original paper by Henkelman.
In the last section of the paper, he described that using the variable spring constants will give a better approximation of the tangent near saddle point. However I cannot find such setups for variable spring constants in CP2K. Has this not been implemented yet or is it hidden somewhere? I also checked the regtest where no similar setup is given.
Thanks for all taking time.