Issues with Configurations using stochastic SHCI

[NastaMauger]

Hello,

While performing SHCI (stochastic) on He with increasing values of sweep_eepsilon (the first value is always kept at 9e-1, and the second value is modified), I aim to obtain the different energies component as a function of the number of determinants. However, with only 2 determinants, I end up with the following configurations:

• Config_1: 200000000 with Coefficient -0.9997315
• Config_2: 000002000 with Coefficient 0.02317060

Should not my second configuration be 020000000 instead? I find these results a bit odd, and when trying to use the deterministic version of DICE, it fails (https://github.com/sanshar/Dice/issues/24).

Here my pyscf script: shci.txt

Thank you for you help