Hiphive is not able to correct the little negative modes

Dear all,

I am mailing you regarding the effect of hiphive on phonon dispersion obtained by Phonopy of a 2D material. I got little negative mode by phonopy (image is attached below). I am trying Hiphive to correct these modes (the code is also attached below). I have tried by increasing the cutoff and changing the parameters (Huang, Born-huang and both), unfortunately none of them are affected these modes. Can you please help me to sort out the issue.
Thanks for your time and consideration.

please see the frequently asked questions from the hiphive documentation here

For example you should use the largest possible 2nd order cutoff and no 3rd order when extracting the parameters. You can also try the multiple different values for alpha as e.g. done here