[lammps-users] integrator

I have a question about the LAMMPS integrator.
I use lj/cut with pair_modify shift yes to do my simulations. Can anyone tell me what kind of equation does it use to shift the potential to zero at the cut-off. Alos, does the first and larger derivatives all go to zero. If this information is listed in the documentaion, can someone send me the link? (I could’nt figure out). I am comparing the results I got using LAMMPS with that of my group mate who uses his own code to do the simulations. We just find differences in the potential energies per atom. Thats why I would like to know the complete form of the potential energy when it is shifted.

All the "shift" does is move the LJ potential up by "delta"
where delta is the value of the potential at the cutoff distance.

Think of adding delta to every pairwise energy if it is within
the cutoff. So it will not go smoothly to 0 (slope = 0) at the
cutoff, unless you use 2^(1/6) as your cutoff.

You can look at the lj/smooth potential, but it has its own issues.