Packing fraction related issues

GROMACS version:2021
GROMACS modification: No
I am running some MD simulations (after energy minimization) for spherical particles by inserting them as point particles into a cubic box (using gmx insert) , with the repulsive part of LJ potential. I am providing all the non-bonded interaction parameters in the param.itp file. If I calculate the packing fraction (particle volume/box volume) of the system by using the sigma(diameter) values that have been provided in param.itp file, it is coming as pf = 0.90. So now I have confused that with this high packing fraction how the system becomes stable (as there is a repulsive part of LJ). Is this system allowing some overlap between the particles (not top to each other)?