Hello,
this is probably a known problem:
We have a problem with questions where the students need to enter a floating point number smaller than 10^(-10). Any answer smaller than 10^(-10) is graded "correct" by STACK.
I used the code abs(sa-ta) < opt in the answer variables, and it delivers correct interpretations.
Attached a screenshot and a question for testing.
Cheers,
Mikko
