Add gradient check to minimization and decrease conditions

[nicholaskl97]

Checklist

• Appropriate tests were added
• Any code changes were done in a way that does not break public API
• All documentation related to code changes were updated
• The new code follows the
  contributor guidelines, in particular the SciML Style Guide and
  COLPRAC.
• Any new documentation only uses public API

Additional context

Add option to train for the fixed point to be a local extremum of both $V$ and $\dot{V}$ by training for their gradients to be zero at the fixed point.